Unmanned aerial vehicle flight parameter prediction reliability evaluation method based on entropy value

Abstract

The invention discloses an entropy-based unmanned aerial vehicle flight parameter prediction reliability evaluation method. The method comprises the following steps: collecting multi-source multi-mode flight data of the same aircraft in the same time period; performing data cleaning and normalization processing on the acquired data by using an auto-encoder network; inputting the clean data into a flight parameter prediction network of a long-short term memory (LSTM) network to obtain a prediction result matrix; the obtained prediction matrix and the original data after dimension reduction are calculated through entropy values, and reliability evaluation is carried out; and finally obtaining the reliability of a prediction result through an entropy calculation result. According to the method, certain reliability evaluation can be given to the prediction result of the flight parameters, the prediction reasonability is judged, then the flight parameters are reasonably adjusted, unreasonable data are avoided, and the method is of great significance to healthy operation of an aircraft.

Description

Technical Field

[0001] The invention belongs to the technical field of unmanned aerial vehicles, and in particular relates to an entropy-based unmanned aerial vehicle flight parameter prediction reliability assessment method. Background Art

[0002] Flight parameters are key data to ensure flight safety, guide flight operations, optimize flight performance, and achieve precise navigation and positioning. Accurate prediction of flight parameters can ensure flight safety, optimize flight paths, reduce energy consumption, and improve flight efficiency. By accurately estimating key parameters such as flight speed, altitude, and heading, pilots and ground control centers can make timely and effective decisions to respond to various emergencies.

[0003] Reliability analysis of predicted flight parameter data is a key step in ensuring flight safety, improving the accuracy of prediction models and the scientific nature of decision-making. It helps to identify and correct deviations and uncertainties in predictions, enhance data confidence, and provide a solid basis for flight control, path planning and risk assessment. Reliability analysis can optimize prediction models, improve flight efficiency, and reduce potential risks, which is very important for ensuring the smooth implementation of flight missions.

[0004] Therefore, it is of great significance to evaluate the reliability of aircraft flight parameter prediction. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a method for evaluating the reliability of UAV flight parameter prediction based on entropy value, which collects multi-source and multi-modal flight data of the same aircraft in the same time period; uses an autoencoder network to clean and normalize the collected data; inputs the clean data into the long short-term memory network (LSTM) flight parameter prediction network to obtain a prediction result matrix; uses entropy value to calculate the obtained prediction matrix and the original data after dimensionality reduction for reliability evaluation; and finally obtains the reliability of the prediction result through the entropy value calculation result. The present invention can give a certain reliability evaluation to the prediction result of the flight parameter, judge the rationality of the prediction, and then reasonably adjust the flight parameters, avoid unreasonable data, which is of great significance to the healthy operation of the aircraft.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] Step 1: Collect multi-source and multi-modal flight data of the same aircraft in the same time period;

[0008] Step 2: For the collected multi-source and multi-modal flight data, the autoencoder network is used to clean and normalize the data to achieve normalization and dimensionality reduction of the multi-source data. The obtained aircraft parameter time matrix is: ,in , Indicates the time period A time series after inner dimension reduction, such a time series has List;

[0009] Step 3: Input the aircraft parameter time matrix into the LSTM flight parameter prediction network to obtain the predicted future time period The flight parameters of the predicted flight parameter matrix are ,in Indicates that it corresponds to The future time period obtained The flight parameters obtained in this way have a total of M columns in the prediction sequence;

[0010] Step 4: Aircraft parameter time matrix Calculate the entropy value according to the time series and get Initial entropy value; the predicted flight parameter matrix obtained in step 3 Time matrix with aircraft parameters Splice into a new matrix ,in , calculate the matrix The entropy value of

[0011] Step 5: Through the Matrix and The entropy value of is used to obtain reliability.

[0012] Furthermore, the normalization process in step 2 is specifically as follows: for the same time period The multi-source multi-modal flight data is used to obtain the maximum value of the time period. , divide the collected raw flight parameter data by its mean value to normalize all the raw flight parameter data, divide the data range into -1 to 1, and obtain the normalized data , Indicates the time period A time series from the same data source, such time series has List.

[0013] Furthermore, the autoencoder network in step 2 has the input , enter The compression and reconstruction process learns an effective data representation and obtains the aircraft parameter time matrix as:

[0014] .

[0015] Furthermore, the objective function of the autoencoder network in step 2 is: ,in Represents the dist function, which calculates the distance between each vector in X and each vector in Y. The returned result Minimizeloss is a matrix of size N×M, where N is the number of vectors in X and M is the number of vectors in Y.

[0016] Furthermore, the LSTM flight parameter prediction network is used to obtain the predicted future time period. The flight parameters of the predicted flight parameter matrix are: , the input is ,The specific structure of the LSTM flight parameter prediction network is as follows:

[0017] The long short-term memory network LSTM consists of three gates: input gate, forget gate and output gate, which are used to control the flow and update of information; there are four sets of weights in the LSTM neuron, namely input gate, output gate, forget gate, and generation of candidate memory; in this network, these four sets of weights are updated through the back propagation algorithm to minimize the prediction error:

[0018] Step 3-1: Input gate determines the extent to which the current input information is written into the memory cell;

[0019] Step 3-2: Forget gate, which determines whether to forget the historical information stored in the memory unit, thereby allowing the long short-term memory network to selectively clear irrelevant information;

[0020] Step 3-3: Output gate determines the output of the information in the memory unit at the current time step.

[0021] Furthermore, the loss function of the LSTM flight parameter prediction network is mean square error.

[0022] Furthermore, the LSTM flight parameter prediction network uses a stochastic gradient descent method to update the weight matrix, optimize the model parameters, and reduce the value of the loss function.

[0023] Furthermore, the LSTM flight parameter prediction network uses permutation entropy to calculate entropy values, and obtains the matrix The length of a column is The entropy value and the new matrix A column length is The entropy value .

[0024] Furthermore, the size of the permutation entropy represents the degree of randomness of the time series. The smaller the value of the permutation entropy, the more regular the time series is. Conversely, the more random the time series is. By statistically analyzing the changes in the degree of disorder of the time series before and after the prediction, it is calculated whether the time series is reasonably predicted, and then the reliability of the flight parameter prediction is inferred.

[0025] Furthermore, the step 5 is specifically as follows:

[0026] Input the flight parameter time matrix X, and get the entropy value sequence of the matrix as: The entropy value sequence obtained after calculating the entropy value of the new time matrix Z obtained in step 4 before and after the prediction is: ; The overall reliability obtained based on the above entropy value sequence is:

[0027] ;

[0028] Step 5-1: By statistically analyzing the entropy values ​​of the time series before and after the prediction, we can determine the change in the disorder degree, calculate whether the time series is reasonably predicted, and then deduce the reliability of the flight parameter prediction; the entropy value is calculated using the permutation entropy, where the calculation method of the permutation entropy is as follows:

[0029] Step 5-1-1: Convert the time series Refactored to Sequence ,in , is the time delay parameter, is the embedding dimension;

[0030] Step 5-1-2: Place each All internal data points are arranged in ascending order and a symbol vector is set Indicates the order of size, the smallest element corresponds to , the largest element corresponds to , if two identical values ​​are generated, they are arranged in order of their subscripts; Include dimension, so Arrangement mode;

[0031] Step 5-1-3: Use symbolic vector representation , at this time each The elements in are expressed in order of their magnitude;

[0032] Step 5-1-4: Calculate the probability of each sequence appearing ;

[0033] Step 5-1-5: According to the form of Shannon entropy, define The permutation entropy of different permutations is as follows:

[0034]

[0035] Step 5-1-6: Calculate the permutation entropy of the data before prediction and obtain a column of permutation entropy values:

[0036] ;

[0037] Step 5-2: The predicted flight parameter matrix is Calculate the permutation entropy and get a column of permutation entropy values .

[0038] The beneficial effects of the present invention are as follows:

[0039] 1. The present invention can give a certain reliability assessment to the prediction results of flight parameters, judge the rationality of the prediction and then reasonably adjust the flight parameters, avoid unreasonable data, which is of great significance to the healthy operation of the aircraft.

[0040] 2. The present invention adopts data dimension reduction technology based on autoencoders, facing complex multivariate and multimodal flight parameters, fully extracting their useful information, using relatively simple data to process complex flight parameter information, and digging deeper data connections. The LSTM prediction network is adopted, facing time series information, avoiding gradient disappearance and gradient explosion, and better predicting time series signals.

[0041] 3. The present invention measures the changes in signal complexity before and after the prediction, measures whether there is a dramatic change in signal complexity, and then infers the reliability of the prediction. The entropy method is simple and rapid, and can calculate the changes in reliability in real time, which helps to achieve real-time update of aircraft parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a flow chart of the method of the present invention;

[0043] Figure 2 Schematic diagram of the autoencoder of the present invention;

[0044] Figure 3 It is a schematic diagram of the long short-term memory neural network of the present invention;

[0045] Figure 4 This is a schematic diagram of the LSTM reliability prediction network of the present invention. DETAILED DESCRIPTION

[0046] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0047] As an important indicator to measure the flight status of an aircraft, flight parameters include axial, lateral and normal overloads, pitch, roll, yaw rate, angle of attack, sideslip angle, flight path angle, path inclination angle, true heading angle, pitch angle, roll angle, true airspeed, ground speed and other aspects. These parameters provide an important reference for the decision-making of pilots and ground control personnel. When the aircraft's navigation system encounters an abnormality or failure, the flight system will transmit erroneous flight data to the flight control system, which poses a serious threat to flight safety.

[0048] Flight parameters are key data to ensure flight safety, guide flight operations, optimize flight performance, and achieve precise navigation and positioning. Accurate prediction of flight parameters can ensure flight safety, optimize flight paths, reduce energy consumption, and improve flight efficiency. By accurately estimating key parameters such as flight speed, altitude, and heading, pilots and ground control centers can make timely and effective decisions to respond to various emergencies.

[0049] Reliability analysis of predicted flight parameter data is a key step in ensuring flight safety, improving the accuracy of prediction models and the scientific nature of decision-making. It helps to identify and correct deviations and uncertainties in predictions, enhance data confidence, and provide a solid basis for flight control, path planning, and risk assessment. Reliability analysis can optimize prediction models, improve flight efficiency, and reduce potential risks, which is of great significance to ensuring the smooth implementation of flight missions.

[0050] Based on the above analysis, the present invention provides a method for measuring the reliability of flight parameter prediction in a short time, such as Figure 1 As shown, the specific steps include:

[0051] Step S1: Collect multi-source and multi-modal flight data of the same aircraft in the same time period;

[0052] Step S2: The collected multi-source multi-modal flight data is cleaned and normalized using an autoencoder network to achieve normalization and dimensionality reduction of the multi-source data. The obtained aircraft parameter time matrix is: ,in , Indicates the time period A time series after inner dimension reduction, such a time series has List;

[0053] Step S3: Input the aircraft parameter time matrix into the LSTM flight parameter prediction network to obtain the predicted future time period The flight parameters of the predicted flight parameter matrix are ,in Indicates that it corresponds to The future time period obtained The flight parameters obtained in this way have a total of M columns in the prediction sequence;

[0054] Step S4: Transform the aircraft parameter time matrix Calculate the entropy value according to the time series and get The initial entropy value; the predicted flight parameter matrix obtained in the previous step Time matrix with aircraft parameters Splice into a new matrix according to the source ,in , calculate the matrix The entropy value of .

[0055] Step S5: Through the matrix and The entropy value of is used to obtain reliability.

[0056] Furthermore, the normalization process in step 2 is specifically as follows: for the same time period The multi-source multi-modal flight data is used to obtain the maximum value of the time period. , divide the collected raw flight parameter data by the mean to normalize all the raw flight parameter data, divide the data range into -1 to 1, and obtain the normalized data , Indicates the time period A time series from the same data source, such a time series has List.

[0057] In step S2, an autoencoder is constructed as Figure 2 As shown, the input is , enter The compression and reconstruction process learns an effective data representation and obtains the aircraft parameter time matrix as:

[0058] .

[0059] S21, building an autoencoder includes the following steps:

[0060] S211, build the encoder;

[0061] S212, build a decoder;

[0062] S213, setting a loss function;

[0063] S214, training.

[0064] In step S2, an autoencoder network is constructed with only one hidden layer. , the activity value of the middle hidden layer of the autoencoder is The encoding of The output of the autoencoder reconstructs the data into

[0065] The objective function of the autoencoder network constructed in step S2 is .

[0066] In step S3, an LSTM flight parameter prediction network is constructed, such as Figure 3 As shown, it is used to get the predicted future time period The flight parameters of the predicted flight parameter matrix are: , the input is ,The specific structure of the LSTM flight parameter prediction network is as follows:

[0067] S31, Long Short-Term Memory Neural Network LSTM mainly contains three key gates: input gate, forget gate and output gate, which are used to control the flow and update of information. The model updates its weights through the back-propagation algorithm to minimize the prediction error.

[0068] S311, set the hidden layer size of the LSTM unit. The larger the value, the stronger the expressive power of the model, but the amount of calculation and the risk of overfitting also increase. Start testing from 50 or 100.

[0069] S312, select the appropriate sequence length. It is usually necessary to select according to the time series characteristics of the data. A sequence that is too short may lead to insufficient information, and a sequence that is too long will increase the computational overhead. According to the actual situation, it is recommended to select more than 1024 data points.

[0070] S313, a smaller batch_size can help the model converge more stably, but the training speed will decrease; a larger batch_size will speed up the training, but may cause model oscillation. Common values ​​are 16, 32, or 64.

[0071] S314, the learning rate is usually set to 0.001 or less, and can be adjusted according to the convergence speed of the model. You can also try to use the learning rate scheduler to dynamically adjust the learning rate.

[0072] S315, you can try stacking multiple layers of LSTM (such as 2-3 layers) to increase the model depth to improve performance, but this will also increase the difficulty of training and the risk of overfitting.

[0073] Step S3 constructs an LSTM flight parameter prediction network, and the loss function is the mean square error.

[0074] Step S3 constructs an LSTM flight parameter prediction network and uses the stochastic gradient descent method to update the weight matrix, optimize the model parameters, and reduce the value of the loss function.

[0075] In step S4, permutation entropy is used to calculate the entropy value. The permutation entropy has the advantages of simple algorithm, fast calculation speed, strong anti-noise ability, etc., and is widely used in the fields of mechanical fault diagnosis and biological signal processing. The length of a column is The entropy value and the new matrix A column length is The entropy value .

[0076] The permutation entropy is used to calculate the matrix before and after the prediction. The size of the permutation entropy represents the randomness of the time series. The smaller the value, the more regular the time series is. Conversely, the more random the time series is. By statistically analyzing the changes in the disorder of the sequence before and after the prediction, it is calculated whether the sequence is reasonably predicted, and then the reliability of the flight parameter prediction is inferred.

[0077] Input the original flight parameter time matrix X, and the entropy value sequence of the matrix is ​​obtained as follows: The entropy value sequence obtained after calculating the entropy value of the new time matrix Z obtained in step 4 before and after the prediction is: The overall reliability obtained based on the above entropy value sequence is:

[0078] ;

[0079] S41, the calculation method of permutation entropy is as follows:

[0080] S411, time series Refactored to Sequence ,in , is the time delay parameter, is the embedding dimension;

[0081] S412, each All internal data points are arranged in ascending order and a symbol vector is set Indicates the order of size, the smallest element corresponds to , the largest element corresponds to , if two identical values ​​are generated, they are arranged in order of their subscripts; Include dimension, so Arrangement mode;

[0082] S413, using symbolic vector representation , at this time each The elements in are expressed in order of their magnitude;

[0083] S414, calculate the probability of each sequence appearing ;

[0084] S415, according to the form of Shannon entropy, define The permutation entropy of different permutations is as follows:

[0085]

[0086] S416, calculating the permutation entropy of the data before prediction, and obtaining a column of permutation entropy values:

[0087] ;

[0088] S42, the predicted flight parameter matrix is Calculate the permutation entropy and get a column of permutation entropy values .

Claims

1. A method for evaluating the reliability of UAV flight parameter prediction based on entropy value, characterized in that: The steps include: Step 1: Collect multi-source and multi-modal flight data of the same aircraft in the same time period; Step 2: For the collected multi-source and multi-modal flight data, the autoencoder network is used to clean and normalize the data to achieve normalization and dimensionality reduction of the multi-source data. The obtained aircraft parameter time matrix is: ,in , Indicates the time period A time series after inner dimension reduction, such a time series has List; Step 3: Input the aircraft parameter time matrix into the LSTM flight parameter prediction network to obtain the predicted future time period The flight parameters of the predicted flight parameter matrix are ,in Indicates that it corresponds to The future time period obtained The flight parameters, the prediction sequence obtained in this way has a total of M columns; Step 4: Aircraft parameter time matrix Calculate the entropy value according to the time series and get An initial entropy value; The predicted flight parameter matrix obtained in step 3 is Time matrix with aircraft parameters Splice into a new matrix ,in , calculate the matrix The entropy value of Step 5: Through the Matrix and The entropy value of is used to obtain reliability.

2. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 1 is characterized in that: The normalization process in step 2 is specifically as follows: The multi-source multi-modal flight data is used to obtain the maximum value of the time period. , divide the collected raw flight parameter data by its mean value to normalize all the raw flight parameter data, divide the data range into -1 to 1, and obtain the normalized data , Indicates the time period A time series from the same data source, such time series has List.

3. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 2 is characterized in that: The autoencoder network in step 2 has the following input: , enter The compression and reconstruction process learns an effective data representation and obtains the aircraft parameter time matrix as: 。 4. The method for predicting reliability of unmanned aerial vehicle flight parameters based on entropy value according to claim 3 is characterized in that: The objective function of the autoencoder network in step 2 is: ,in Represents the dist function, which calculates the distance between each vector in X and each vector in Y. The returned result Minimizeloss is a matrix of size N×M, where N is the number of vectors in X and M is the number of vectors in Y.

5. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 4 is characterized in that: The LSTM flight parameter prediction network is used to obtain the predicted future time period The flight parameters of the predicted flight parameter matrix are: , the input is ,The specific structure of the LSTM flight parameter prediction network is as follows: The long short-term memory network LSTM consists of three gates: input gate, forget gate and output gate, which are used to control the flow and update of information; there are four sets of weights in the LSTM neuron, namely input gate, output gate, forget gate, and generation of candidate memory; in this network, these four sets of weights are updated through the back propagation algorithm to minimize the prediction error: Step 3-1: Input gate determines the extent to which the current input information is written into the memory cell; Step 3-2: Forget gate, which determines whether to forget the historical information stored in the memory unit, thereby allowing the long short-term memory network to selectively clear irrelevant information; Step 3-3: Output gate determines the output of the information in the memory unit at the current time step.

6. The method for predicting reliability of unmanned aerial vehicle flight parameters based on entropy value according to claim 5 is characterized in that: The loss function of the LSTM flight parameter prediction network is the mean square error.

7. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 6 is characterized in that: The LSTM flight parameter prediction network uses the stochastic gradient descent method to update the weight matrix, optimize the model parameters, and reduce the value of the loss function.

8. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 7 is characterized in that: The LSTM flight parameter prediction network uses permutation entropy to calculate entropy values ​​and obtains the matrix The length of a column is The entropy value and the new matrix A column length is The entropy value .

9. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 8 is characterized in that: The size of the permutation entropy represents the randomness of the time series. The smaller the value of the permutation entropy, the more regular the time series is. Conversely, the more random the time series is. By statistically predicting the change in the disorder of the time series before and after the prediction, it is calculated whether the time series is reasonably predicted, and then the reliability of the flight parameter prediction is inferred.

10. The method for predicting reliability of UAV flight parameters based on entropy value according to claim 9 is characterized in that: The step 5 is specifically as follows: Input the flight parameter time matrix X, and get the entropy value sequence of the matrix as: The entropy value sequence obtained after calculating the entropy value of the new time matrix Z obtained in step 4 before and after the prediction is: ; The overall reliability obtained based on the above entropy value sequence is: ； Step 5-1: By statistically analyzing the entropy values ​​of the time series before and after the prediction, we can determine the change in the disorder degree, calculate whether the time series is reasonably predicted, and then deduce the reliability of the flight parameter prediction; the entropy value is calculated using the permutation entropy, where the calculation method of the permutation entropy is as follows: Step 5-1-1: Convert the time series Refactored to Sequence ,in , is the time delay parameter, is the embedding dimension; Step 5-1-2: Place each All internal data points are arranged in ascending order and a symbol vector is set Indicates the order of size, the smallest element corresponds to , the largest element corresponds to , if two identical values ​​are generated, they are arranged in order of their subscripts; Include dimension, so Arrangement mode; Step 5-1-3: Use symbolic vector representation , at this time each The elements in are expressed in order of their magnitude; Step 5-1-4: Calculate the probability of each sequence appearing ; Step 5-1-5: According to the form of Shannon entropy, define The permutation entropy of different permutations is as follows: ； Step 5-1-6: Calculate the permutation entropy of the data before prediction and obtain a column of permutation entropy values: ； Step 5-2: The predicted flight parameter matrix is Calculate the permutation entropy and get a column of permutation entropy values .

Images