A method for quickly matching launch parameters of a rocket-assisted unmanned aerial vehicle

Abstract

The present application relates to a kind of unmanned aerial vehicle rocket boost launch parameter quick matching method, comprising: establishing launch parameter set, the uncontrollable parameter in launch parameter set is regarded as the input parameter of quick matching, controllable parameter is regarded as the output parameter of quick matching;Establish unmanned aerial vehicle expected launch safety value evaluation model;Obtain launch test sample library by simulation experiment method, each test sample in library contains launch parameter set and its corresponding expected launch safety value;The expected launch safety value is graded, and the test sample that meets the safety level requirement is screened from launch test sample library;The test sample screened is used to train, check quick matching model, so that quick matching model can be according to the uncontrollable launch parameter input, output reaches the controllable launch parameter of expected launch safety value level;Unmanned aerial vehicle launch parameter quick matching can be carried out using quick matching model.The present application can realize the quick matching of unmanned aerial vehicle rocket boost launch parameter.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) rocket booster launch technology, and specifically to a method for rapid matching of UAV rocket booster launch parameters. Background Technology

[0002] Rocket boosting is a common launch method for drones. It has low requirements for site and environment, is easy to deploy and retrieval, and can meet the needs of drones for rapid and mobile launch.

[0003] During launch, the configuration, aerodynamic and weight characteristics, and flight parameters of the UAV undergo drastic nonlinear changes, posing a significant challenge to its launch safety. Numerous factors, varying in degree, influence UAV launch safety, including atmospheric conditions such as temperature and wind, booster rocket installation angle deviations, aerodynamic coupling effects between the rocket and the UAV, and launch phase control strategies. Therefore, only reasonable and reliable selection and matching of launch parameters can reduce launch risks and improve launch safety margins.

[0004] Currently, the main methods for designing and verifying launch parameters for rocket-assisted unmanned aerial vehicles (UAVs) are computer simulation and ignition test firing. Of these two methods, ignition test firing requires the use of a dummy with parameters close to those of a real aircraft or the direct use of a real aircraft, resulting in high testing costs, limited verification capabilities, and the ability to verify launch parameters only under a specific set of conditions. Therefore, computer simulation technology is still primarily used to build relatively accurate aircraft launch dynamics and control models for tasks such as launch parameter safety boundary assessment, launch parameter matching design, and the dynamic impact of external interference on the launch process. This approach has a certain degree of engineering practicality and acceptance.

[0005] The existing method for matching launch parameters mainly involves pre-discretizing several launch parameters based on a simulation model to form a series of initial launch parameter sets. By running simulations and analyzing the flight parameter response during launch, the method evaluates whether each set of launch parameters meets safety requirements, thus establishing a set of optional launch parameters. Given launch conditions (launch altitude, ambient temperature, wind speed, and wind direction, etc.), the optimal set of launch parameters is selected from this set for use. However, this method has several drawbacks: 1. The set of optional launch parameters is discrete, and the selected matching values ​​may not achieve the optimal launch state; 2. If launch conditions change temporarily, the re-matching operation is complex, time-consuming, and may exceed the range of optional launch parameters; 3. To improve model accuracy, connecting to fixed equipment to form a semi-physical simulation system increases model complexity and limits portability and transferability, making it difficult to deploy and use conveniently during field test flights. Summary of the Invention

[0006] Based on the above analysis, the present invention aims to provide a method for rapid matching of rocket booster launch parameters for unmanned aerial vehicles (UAVs) to solve the problem of rapid matching of rocket booster launch parameters for UAVs.

[0007] This invention discloses a method for rapid matching of rocket booster launch parameters for unmanned aerial vehicles (UAVs), comprising:

[0008] Establish a set of launch parameters, including environmental parameters, ground-adjustable parameters, and launch control parameters; use the uncontrollable parameters in the launch parameter set as input parameters for fast matching, and the controllable parameters as output parameters for fast matching;

[0009] Establish a model for assessing expected launch safety values, and determine the expected launch safety values ​​based on launch safety assessment indicators;

[0010] A launch test sample library was obtained through simulation experiments. Each launch test sample in the library includes a set of launch parameters and the expected launch safety value corresponding to that set of launch parameters.

[0011] The expected launch safety value range is classified, and test samples that meet the safety level requirements are selected from the launch test sample library;

[0012] The selected test samples are used to train and verify the fast matching model, so that the fast matching model can output controllable launch parameters that achieve the desired launch safety level based on the input uncontrollable launch parameters.

[0013] Rapid matching of UAV launch parameters is achieved using a fast matching model.

[0014] Furthermore, the uncontrollable parameters in the launch parameter set include: ambient temperature, launch altitude, wind speed and direction, launch conditions, and the rocket's longitudinal and lateral installation angles;

[0015] The controllable parameters in the launch parameter set include: launch angle and launch longitudinal and lateral control parameters.

[0016] Furthermore, the process of establishing the expected launch security value assessment model includes:

[0017] 1) Establish an evaluation index system for expected launch safety values;

[0018] 2) Standardize the evaluation indicators;

[0019] 3) Determine the weight values ​​of the evaluation indicators;

[0020] 4) Calculate the expected launch safety value by combining the actual values ​​and weight values ​​of the evaluation indicators.

[0021] Furthermore, the evaluation indicators in the expected launch safety value assessment index system include the maximum launch body pitch angle θ. f.maxMinimum angle of attack α of the launcher f.min At the end of the boost, the flight altitude h f At the end of the boost, the flight speed V f Peak roll angle φ of the transmitter body f.max and the peak sideslip angle β of the transmitter body f.max ;

[0022] The standardized evaluation index vector obtained by standardizing the evaluation indicators for:

[0023]

[0024] in, and These are the maximum pitch angle of the launcher, the minimum angle of attack of the launcher, the flight altitude at the end of the boost, the flight speed at the end of the boost, the peak roll angle of the launcher, and the peak sideslip angle of the launcher, respectively, after standardization.

[0025] The index weight value vector obtained by determining the weight values ​​of the evaluation indicators for:

[0026]

[0027] in, and These are the index weights corresponding to the maximum launcher pitch angle, minimum launcher angle of attack, flight altitude at the end of boost, flight speed at the end of boost, peak launcher roll angle, and peak launcher sideslip angle, respectively; T represents vector transpose;

[0028] Expected launch safety value

[0029] Furthermore, a five-level evaluation criterion is adopted, and a first evaluation threshold S is set in descending order of the expected launch safety value range [0.0, 1.0]. d1 Second evaluation threshold S d2 The third evaluation threshold S d3 and the fourth evaluation threshold S d4 ;

[0030] When classifying the expected range of safe launch values,

[0031] Determine the expected launch safety range (S) d1 [1.0], classified as "Excellent" in evaluation level;

[0032] Determine the expected launch safety range (S) d2 ,S d1 The rating is classified as "Good".

[0033] Determine the expected launch safety range (S) d3 ,S d2 The evaluation level is classified as "Medium".

[0034] Determine the expected launch safety range (S) d4 ,S d3 ], which is classified as "poor" in evaluation;

[0035] The expected launch safety value range is (0.0, S). d4 ], classified as "inferior" in evaluation level;

[0036] When screening test samples, test samples with the expected launch safety value of "excellent" are selected and retained.

[0037] Furthermore, when obtaining the launch test sample library through simulation experiments, launch test samples are generated by sampling from the UAV launch parameter set using the Latin hypersolution method or the Monte Carlo method.

[0038] Furthermore, the fast matching model is a BP neural network model; the input of the BP neural network model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

[0039] Furthermore, in the design process of the fast matching model, the selected test samples are divided into training group and verification group, which are used to complete the model training and error calibration respectively; and the model training and verification are repeated through multiple rounds of cross-grouping to improve the model matching quality.

[0040] Furthermore, the fast matching model is a polynomial response surface model; the input of the polynomial response surface model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

[0041] Furthermore, in the design process of the fast matching model, the selected test samples are divided into training group and verification group, which are used to complete the model training and error calibration respectively; and the model training and verification are repeated through multiple rounds of cross-grouping to improve the model matching quality.

[0042] This invention can achieve one of the following beneficial effects:

[0043] The method for rapid matching of launch parameters for UAV rocket boosters disclosed in this invention has a simple model structure, high matching efficiency, and is easy to deploy and use in various field test flight environments. At the same time, it can quickly respond to and match the optimal aircraft launch parameters when launch conditions change temporarily. Attached Figure Description

[0044] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0045] Figure 1 This is a flowchart of the rapid matching method for drone rocket booster launch parameters in an embodiment of the present invention;

[0046] Figure 2 This is an example diagram illustrating the generation of launch test samples using existing UAV launch dynamics and control models in an embodiment of the present invention.

[0047] Figure 3 This is a schematic diagram of the five-level evaluation criterion for uniform expected launch safety value in an embodiment of the present invention;

[0048] Figure 4 This is a schematic diagram of the BP neural network model structure in an embodiment of the present invention. Detailed Implementation

[0049] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and, together with the embodiments of the present invention, serve to illustrate the principles of the present invention.

[0050] One embodiment of the present invention discloses a method for rapid matching of rocket booster launch parameters for unmanned aerial vehicles, such as... Figure 1 As shown, it includes:

[0051] Step S1: Establish a set of launch parameters including environmental parameters, ground adjustable parameters, and launch control parameters; use the uncontrollable parameters in the launch parameter set as input parameters for fast matching, and the controllable parameters as output parameters for fast matching;

[0052] Step S2: Establish a desired launch safety value assessment model and determine the desired launch safety value based on the launch safety assessment indicators;

[0053] Step S3: Obtain a launch test sample library through simulation experiments; each launch test sample in the library includes a set of launch parameters and the expected launch safety value corresponding to that set of launch parameters;

[0054] Step S4: Classify the expected launch safety value range and select test samples that meet the safety level requirements from the launch test sample library;

[0055] Step S5: Train and verify the fast matching model using the selected test samples so that the fast matching model can output controllable launch parameters that achieve the desired launch safety level based on the input uncontrollable launch parameters.

[0056] Step S6: Use the fast matching model to quickly match the launch parameters of the UAV.

[0057] Specifically, in step S1, the set of launch parameters, including environmental parameters, ground adjustment parameters, and launch control parameters, includes environmental parameters which are uncontrollable and may undergo temporary changes before the UAV launch; ground adjustment parameters can be adjusted through experiments or by adjusting the values ​​of ground equipment to change the launch state of the aircraft, but the adjustment of some parameters is time-consuming and difficult to operate on-site; launch control parameters are related to the launch control model architecture and the control law used, and can be easily adjusted before launch.

[0058] In an example of a common launch control method for fixed-wing UAVs, the pitch control loop feeds back the pitch rate and pitch angle to the elevator. Pitch rate feedback increases system pitch damping, while pitch angle feedback controls stable pitch attitude, thereby improving longitudinal performance characteristics. Similarly, the roll control loop feeds back the roll rate and roll angle to the ailerons. Roll rate feedback increases roll damping, while roll angle feedback controls stable roll attitude. Yaw rate is fed back to the rudder to increase directional damping.

[0059] The control law is as follows:

[0060]

[0061] Where, δ t For throttle command, δ t.max The maximum throttle opening; the five launch control parameters are: pitch angle proportionality coefficient k. θ Pitch rate proportional coefficient k q Roll angle proportionality coefficient k φ Roll rate proportionality coefficient k p yaw rate proportionality coefficient k r .

[0062] The set of launch parameters for a UAV is represented as Ω={···,λ i ,···}(i=1,...,n), where n is the number of emission parameters in the set, λ i Let be a certain launch parameter; a definite set of launch parameters can completely determine the rocket-assisted launch process of the UAV; the range of each launch parameter is given simultaneously, and can be expressed as λ. i ∈[λ i.min ,λ i.max ].

[0063] Each specific launch parameter in the launch parameter set can be grouped into controllable and uncontrollable categories. Uncontrollable parameters in the launch parameter set are used as input parameters for rapid matching, while controllable parameters are used as output parameters for rapid matching. Among these, environmental parameters are uncontrollable, and parameters in the ground adjustment parameters, including the longitudinal and lateral installation angles of the rocket boosters, are difficult to adjust on-site. In this embodiment, these are all grouped into uncontrollable parameters. On the other hand, parameters in the ground adjustment parameters that are easy to adjust, as well as launch control parameters, can be conveniently adjusted on-site to improve the safety of the UAV launch process as much as possible. In this embodiment, these are all grouped into controllable parameters.

[0064] To improve the safety of the UAV launch process, in a more specific scheme, the parameters within the launch parameter set are categorized as follows:

[0065] Uncontrollable parameters in the launch parameter set include: ambient temperature, launch altitude, wind speed and direction, launch conditions, and the rocket's longitudinal and lateral installation angles;

[0066] The controllable parameters in the launch parameter set include: launch angle and launch longitudinal and lateral control parameters.

[0067] Furthermore, uncontrollable parameters in the launch parameter set are used as input parameters for rapid matching, and controllable parameters are used as output parameters for rapid matching; thereby enabling rapid matching and adjustment of controllable parameters in the launch parameters to improve the safety of the UAV launch process.

[0068] Specifically, in step S2, the process of establishing the expected launch security value assessment model includes:

[0069] 1) Establish an evaluation index system for expected launch safety values;

[0070] By integrating expert resources in the field involved in this application, an expert system is established, and an evaluation index system for expected launch safety values ​​is constructed through expert questioning.

[0071] The evaluation indicators in the expected launch safety value assessment index system include the maximum launch body pitch angle θ. f.max Minimum angle of attack α of the launcher f.min At the end of the boost, the flight altitude h f At the end of the boost, the flight speed V f Peak roll angle of the transmitter |φ| f.max , peak sideslip angle of the transmitter body |β| f.max ;

[0072] The evaluation indicators in the expected launch safety value evaluation index system are combined into an evaluation index vector.

[0073] 2) Standardize the evaluation indicators;

[0074] Specifically, the standardized evaluation indicator vector is obtained by standardizing the evaluation indicators. for:

[0075]

[0076] in, and These are the maximum pitch angle of the launcher, the minimum angle of attack of the launcher, the flight altitude at the end of the boost, the flight speed at the end of the boost, the peak roll angle of the launcher, and the peak sideslip angle of the launcher, respectively, after standardization.

[0077] Optionally, existing standardization methods can be used when standardizing evaluation indicators.

[0078] 3) Determine the weight values ​​of the evaluation indicators;

[0079] The expert system can continue to be used, and the extended Bayesian method can be adopted to integrate the judgment information of multiple experts, assign weight values ​​to each evaluation indicator in the evaluation indicator vector, and generate an evaluation indicator weight value vector.

[0080] Specifically, the indicator weight value vector is obtained by determining the weight values ​​of the evaluation indicators. for:

[0081] in, and These are the index weight values ​​corresponding to the maximum pitch angle of the launcher, the minimum angle of attack of the launcher, the flight altitude at the end of the boost, the flight speed at the end of the boost, the peak roll angle of the launcher, and the peak sideslip angle of the launcher, respectively.

[0082] 4) Calculate the expected launch safety value by combining the actual values ​​and weight values ​​of the evaluation indicators;

[0083] The method for calculating the expected launch safety value of a drone is as follows: Its value ranges between 0.0 and 1.0. The larger the value, the safer the drone launch process and the lower the launch risk.

[0084] Specifically, step S3 involves obtaining a launch test sample library through simulation experiments. The simulation experiments are based on the established UAV launch dynamics and control simulation (digital or hardware-in-the-loop) model, and a large number of launch test samples are generated through the Design of Experiments (DOE) method to form the launch test sample library.

[0085] The established UAV launch dynamics and control simulation model is a simulation model that takes the set of launch parameters as input and takes the evaluation indicators, including the evaluation index system of the expected launch safety value, as output.

[0086] Furthermore, in this embodiment, the output of the UAV launch dynamics and control simulation model is input into the expected launch safety value evaluation model established in step S2 to evaluate the expected launch safety value and obtain the expected launch safety value.

[0087] The set of launch parameters for each input UAV launch dynamics and control simulation model, along with the expected launch safety value corresponding to that set of launch parameters, is used as a launch test sample in the launch test sample library.

[0088] Specifically, such as Figure 2 As shown, an example of generating launch test samples using existing UAV launch dynamics and control models is presented.

[0089] exist Figure 2 First, a set of launch parameters, including environmental parameters, ground adjustment parameters, and launch control parameters, is input into the UAV launch dynamics and control model. This model includes a launch control model and a launch dynamics model. The ground adjustment parameters, launch control parameters, and flight parameters fed back from the launch dynamics model are input into the launch control model to calculate the control parameters. These control parameters, along with the environmental and ground adjustment parameters, are then input into the launch dynamics model to generate the flight parameters for the next state. These parameters are fed back to the reflection control model and output to the desired launch safety value assessment model F for safety value assessment, thus obtaining the desired launch safety value. This allows the model to generate an assessment value corresponding to a given set of launch parameters.

[0090] in, Figure 2 In the middle, the flight parameters include the maximum pitch angle θ of the launcher. f.max Minimum angle of attack α of the launcher f.min At the end of the boost, the flight altitude h f At the end of the boost, the flight speed V f Peak roll angle of the transmitter |φ| f.max And the peak value of the sideslip angle of the transmitter body |β| f.max The expected launch safety value assessment model F is the expected launch safety value assessment model established in step S2.

[0091] More specifically, Figure 2 In this context, the UAV launch dynamics and control model can adopt existing models, such as the publicly available UAV launch dynamics and control model in the literature "Research on Modeling and Simulation of Zero-Length Launch with Rocket Boost". This model can perform UAV launch dynamics and control simulation with launch parameters, including the launch parameter set in this embodiment, as input, and output flight parameters, including the launch safety assessment index in this embodiment.

[0092] Optionally, when obtaining the launch test sample library through simulation experiments, launch test samples are generated by sampling from the UAV launch parameter set using the Latin hypercube method.

[0093] For example, when the number of experimental sampling strata for each launch parameter in the launch parameter set are {···,c i When i = 1, ..., n, the total number of experimental samples obtained by sampling using the Latin hypercube method is calculated as follows: c i Let n be the number of layers in the i-th layer, and n be the number of layers.

[0094] In an alternative approach, when obtaining the launch test sample library through simulation experiments, launch test samples are generated by sampling from the UAV launch parameter set using the Monte Carlo method.

[0095] Specifically, in step S4, in order to obtain high-quality samples for training the fast matching model, in this embodiment, the expected emission security value of each sample is classified, and samples with higher security levels are selected for training the fast matching model based on the classification results.

[0096] In the preferred embodiment, a five-level evaluation criterion is adopted, and a first evaluation threshold S is set in descending order of the desired launch safety value range [0.0, 1.0]. d1 Second evaluation threshold S d2 The third evaluation threshold S d3 and the fourth evaluation threshold S d4 ;

[0097] When classifying the expected range of safe launch values,

[0098] Determine the expected launch safety range (S) d1 [1.0], classified as "Excellent" in evaluation level;

[0099] Determine the expected launch safety range (S) d2 ,S d1 The rating is classified as "Good".

[0100] Determine the expected launch safety range (S) d3 ,S d2 The evaluation level is classified as "Medium".

[0101] Determine the expected launch safety range (S) d4 ,S d3 ], which is classified as "poor" in evaluation;

[0102] The expected launch safety value range is (0.0, S). d4 ], classified as "inferior" in evaluation level;

[0103] When screening test samples, test samples with the expected launch safety value of "excellent" are selected and retained.

[0104] like Figure 3 In the example shown, the first evaluation threshold S is given. d1 =0.8, Second evaluation threshold S d2 =0.6, Third evaluation threshold S d3 =0.4 and the fourth evaluation threshold S d4 A schematic diagram illustrating the classification of expected launch safety levels using a five-level evaluation criterion of 0.2.

[0105] Specifically, in step S5, the fast matching model is a BP neural network model; the input of the BP neural network model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

[0106] Specific BP neural network models, such as Figure 4 The diagram shows a single-hidden-layer BP neural network model. The input parameters are ambient temperature, launch altitude, wind speed and direction, launch conditions, and the rocket's longitudinal and lateral installation angles. The output parameters are the launch angle and the launch longitudinal and lateral control parameters. The number of neurons in the input and output layers of the BP neural network model matches the number of input and output parameters, respectively.

[0107] The number of neurons in the hidden layer is set according to the following formula:

[0108]

[0109] Where k is the number of neurons in the hidden layer, m is the number of neurons in the input layer, n is the number of neurons in the output layer, and a is a random integer between 1 and 10.

[0110] In the specific fast matching model training process, the selected test samples are divided into training group and verification group, which are used to complete the model training and error calibration respectively; and the model training and verification are repeated through multiple rounds of cross-grouping to improve the model matching quality.

[0111] More specifically, in practice, multiple fast matching models with different structures can be built by taking different values ​​for 'a'. By repeating the "training-verification" process of the models, the value of 'a' with the smallest verification error is selected for the structural design of the fast matching model, so as to meet the needs of fast matching and reduce the size of the neural network.

[0112] In an alternative approach, in step S5, the fast matching model is a polynomial response surface model; the input of the polynomial response surface model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

[0113] The specific polynomial response surface model can adopt the polynomial response surface model structure in existing public literature, so that its input layer and output layer are matched with the input parameters and output parameters.

[0114] Furthermore, similar to the training of BP neural network models, the training process of multinomial response surface models involves grouping the selected experimental samples into training and calibration groups to complete the training and error calibration of the model for rapid matching; and improving the model quality by repeating the grouping process multiple times.

[0115] Specifically, in step S6, the fast matching model, after completing model training and verification, can be separated from the UAV launch simulation model in the simulation experiment and run offline independently. By inputting a given set of launch parameters into the fast matching model, the launch parameters with a high expected launch safety level can be calculated and output.

[0116] In summary, the rapid matching method for UAV rocket booster launch parameters in this embodiment has a simple model structure, high matching efficiency, and is easy to deploy and use in various field test flight environments. At the same time, it can quickly respond to and match the optimal aircraft launch parameters when launch conditions change temporarily.

[0117] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for rapid matching of rocket booster launch parameters for unmanned aerial vehicles (UAVs), characterized in that, include: Establish a set of launch parameters, including environmental parameters, ground-adjustable parameters, and launch control parameters; Uncontrollable parameters in the set of transmission parameters are used as input parameters for fast matching, and controllable parameters are used as output parameters for fast matching. Establish a model for assessing expected launch safety values, and determine the expected launch safety values ​​based on launch safety assessment indicators; A sample library for launch tests was obtained through simulation experiments. Each launch test sample in the library includes a set of launch parameters and the expected launch safety value corresponding to that set of launch parameters; Among them, the simulation experiments are based on the established UAV launch dynamics and control simulation model, and launch test samples are generated through experimental design methods to form a launch test sample library; The expected launch safety value range is classified, and test samples that meet the safety level requirements are selected from the launch test sample library; The selected test samples are used to train and verify the fast matching model, enabling the fast matching model to output controllable launch parameters that achieve the desired launch safety level based on the input uncontrollable launch parameters; the fast matching model is used for rapid matching of UAV launch parameters.

2. The method for rapid matching of UAV rocket booster launch parameters according to claim 1, characterized in that, Uncontrollable parameters in the launch parameter set include: ambient temperature, launch altitude, wind speed and direction, launch conditions, and the rocket's longitudinal and lateral installation angles; The controllable parameters in the launch parameter set include: launch angle and launch longitudinal and lateral control parameters.

3. The method for rapid matching of UAV rocket booster launch parameters according to claim 1, characterized in that, The process of establishing the expected launch security value assessment model includes: 1) Establish an evaluation index system for expected launch safety values; 2) Standardize the evaluation indicators; 3) Determine the weight values ​​of the evaluation indicators; 4) Calculate the expected launch safety value by combining the actual values ​​and weight values ​​of the evaluation indicators.

4. The method for rapid matching of UAV rocket booster launch parameters according to claim 3, characterized in that, The evaluation indicators in the expected launch safety value assessment index system include the maximum launch vehicle pitch angle. Minimum angle of attack of the launcher Flight altitude at the end of boost Flight speed at the end of boost peak roll angle of the launcher and the peak sideslip angle of the transmitter body ; The standardized evaluation index vector obtained by standardizing the evaluation indicators for: ； in, , , , , and These are the maximum pitch angle of the launcher, the minimum angle of attack of the launcher, the flight altitude at the end of the boost, the flight speed at the end of the boost, the peak roll angle of the launcher, and the peak sideslip angle of the launcher, respectively, after standardization. The index weight value vector obtained by determining the weight values ​​of the evaluation indicators for: ； in, , , , and These are the index weight values ​​corresponding to the maximum pitch angle of the launcher, the minimum angle of attack of the launcher, the flight altitude at the end of the boost, the flight speed at the end of the boost, the peak roll angle of the launcher, and the peak sideslip angle of the launcher, respectively. Indicates vector transpose; Expected launch safety value , .

5. The method for rapid matching of UAV rocket booster launch parameters according to claim 4, characterized in that, A five-level evaluation criterion is adopted, within the range of expected launch safety values. The first evaluation threshold is set by sorting the values ​​from largest to smallest. Second evaluation threshold Third evaluation threshold and the fourth evaluation threshold ; When classifying the expected range of safe launch values, Desired launch safety range It is classified as "Excellent" in evaluation. Desired launch safety range It is classified as "Good" in terms of evaluation level; Desired launch safety range It is classified as "medium" in the evaluation level; Desired launch safety range It is classified as "poor" in evaluation. Desired launch safety range It is classified as "inferior" in the evaluation level; When screening test samples, test samples with the expected launch safety value of "excellent" are selected and retained.

6. The method for rapid matching of UAV rocket booster launch parameters according to any one of claims 1-5, characterized in that, When obtaining a launch test sample library through simulation experiments, launch test samples are generated by sampling from the UAV launch parameter set using the Latin hypersolution method or the Monte Carlo method.

7. The method for rapid matching of UAV rocket booster launch parameters according to any one of claims 1-5, characterized in that, The fast matching model is a BP neural network model; the input of the BP neural network model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

8. The method for rapid matching of UAV rocket booster launch parameters according to claim 7, characterized in that, In the design process of the fast matching model, the selected test samples are divided into training group and verification group, which are used to complete the model training and error calibration respectively; and the model training and verification are repeated through multiple rounds of cross-grouping to improve the model matching quality.

9. The method for rapid matching of UAV rocket booster launch parameters according to any one of claims 1-5, characterized in that, The fast matching model is a polynomial response surface model; the input of the polynomial response surface model is the uncontrollable parameters in the emission parameter set, and the output is the controllable parameters in the emission parameter set.

10. The method for rapid matching of UAV rocket booster launch parameters according to claim 9, characterized in that, In the design process of the fast matching model, the selected test samples are divided into training group and verification group, which are used to complete the model training and error calibration respectively; and the model training and verification are repeated through multiple rounds of cross-grouping to improve the model matching quality.

Images