CN115480582A - Maneuver prediction method for LSTM-based target, electronic device and storage medium - Google Patents
Maneuver prediction method for LSTM-based target, electronic device and storage medium Download PDFInfo
- Publication number
- CN115480582A CN115480582A CN202210986452.6A CN202210986452A CN115480582A CN 115480582 A CN115480582 A CN 115480582A CN 202210986452 A CN202210986452 A CN 202210986452A CN 115480582 A CN115480582 A CN 115480582A
- Authority
- CN
- China
- Prior art keywords
- target
- maneuver
- unmanned aerial
- aerial vehicle
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 44
- 238000012360 testing method Methods 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 230000001133 acceleration Effects 0.000 claims description 57
- 230000009471 action Effects 0.000 claims description 22
- 230000006870 function Effects 0.000 claims description 18
- 238000002474 experimental method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000004088 simulation Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 230000036544 posture Effects 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 2
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- RZVHIXYEVGDQDX-UHFFFAOYSA-N 9,10-anthraquinone Chemical compound C1=CC=C2C(=O)C3=CC=CC=C3C(=O)C2=C1 RZVHIXYEVGDQDX-UHFFFAOYSA-N 0.000 description 1
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
- G05D1/0808—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a maneuvering prediction method, electronic equipment and storage medium of an LSTM-based target, comprising the following steps of; step 1: constructing a target aircraft maneuvering selection data set in the unmanned aerial vehicle fighting process; step 2: extracting space situation characteristics of a target maneuver countermeasure process; and step 3: segmenting and preprocessing a data set; and 4, step 4: establishing a maneuvering prediction network model of the target based on the LSTM; and 5: using training sets
Training a maneuver prediction network model for an LSTM-based target and utilizing a test set
And (6) carrying out accuracy detection. The invention can predict single maneuvering control parameter or all maneuvering control parametersThe problem of correct prediction can be well solved.
Description
Technical Field
The invention relates to the technical field of flight control and artificial intelligence of aircrafts, in particular to a maneuvering prediction method of an LSTM-based target, electronic equipment and a storage medium.
Background
In recent years, along with the rapid development of the relevant technologies of unmanned aerial vehicles, the application of unmanned aerial vehicles in the military field is more and more extensive, and people pay attention to the problem of predicting the next-stage maneuvering action of a target unmanned aerial vehicle. In the air combat countermeasure process, the unmanned aerial vehicles of both sides carry out the maneuver action of strategy selection next stage as the intelligent agent to let own unmanned aerial vehicle be in the advantage state in the space occupation, and how to foresee the maneuver action that target unmanned aerial vehicle next stage will take in advance is the key problem of own maneuver action selection strategy.
At present, the method for predicting the maneuver of the target unmanned aerial vehicle regards the maneuver selection of the target unmanned aerial vehicle as a fixed weighted sum of several maneuvers, and the weight of each maneuver is determined by expert knowledge. However, the unmanned aerial vehicle maneuver selection is a continuous and dynamic space situation occupation optimizing process, the fixed maneuver weighted sum prediction method does not consider the space situation characteristics of the target maneuver selection, neglects the dynamic and time sequence characteristics of the target maneuver, and does not consider the maneuver perception condition of the target machine on the unmanned aerial vehicle of the party, so that the rule of the target maneuver selection is difficult to accurately describe, and the future maneuver prediction of the target machine is inaccurate.
Disclosure of Invention
To overcome the drawbacks of the above techniques, it is an object of the present invention to provide a maneuver prediction method, an electronic device and a storage medium for an LSTM-based target, which can achieve good results for both the prediction of a single maneuver control parameter and the simultaneous prediction of correct problems for all maneuver control parameters.
In order to achieve the purpose, the invention adopts the technical scheme that:
a maneuver prediction method for an LSTM-based target includes the following steps;
step 1: constructing a target maneuver selection data set S in the unmanned aerial vehicle fighting process, wherein the target refers to a target unmanned aerial vehicle which is in fighting with the unmanned aerial vehicle of one party;
and 2, step: extracting space situation characteristics of a target maneuver countermeasure process;
and step 3: segmenting and preprocessing the target mobile selection data set S in the step 1;
and 4, step 4: establishing a maneuvering prediction network model of the target based on the LSTM;
and 5: using training sets
Training a maneuver prediction network model for an LSTM-based target and utilizing a test set
And (6) carrying out accuracy detection.
Selecting 280 groups of air combat countermeasure data as a data set S of target maneuver prediction, and at any time t, obtaining state information S of unmanned aerial vehicles of both parties of the battle t And motor controlInformation c t Each contains the following information in multiple dimensions, as shown in the formula:
s t ＝[X t ,Y t ,Z t ,EX t ,EY t ,EZ t ,AX t ,AY t ,AZ t ,V t ]
c t ＝[a t ,p t ,y t ]
in the formula, X t ,Y t ,Z t Respectively representing the three-dimensional spatial position coordinates, EX, of the drone at time t t ,EY t ,EZ t Respectively represent the attitude Euler Angle (AX) of the unmanned plane at the moment t t ,AY t ,AZ t ) Form the aircraft nose orientation vector at moment t of unmanned aerial vehicle, V t Speed information, s, representing the unmanned plane at time t t Containing status information of 10 dimensions, a t ,p t ,y t Acceleration information, pitch angle information and yaw angle information, c, representing the speed and direction of the controlling drone, respectively t Contains maneuvering control information of 3 dimensions, and is used for state information of unmanned aerial vehicle at any time t
For indicating, maneuvering control information
Represents; state information of target unmanned aerial vehicle
For indicating, maneuvering control information
Represents; data set S = [ S ] u ,c u ,s e ,c e ]。
In the step 2, in the process of short-distance air combat confrontation, the key of the unmanned aerial vehicle winning is that when the target machine is in the attack range of the own unmanned aerial vehicle, the own unmanned aerial vehicle is in a dominant state in terms of angle situation, the value of the azimuth angle p of the target machine and the value of the entrance angle q of the target machine approach to 0, and the target machine is subjected to the win-win situationWhen the maneuver is predicted, the distance d between the target machine and the angle situation T of the target machine need to be considered a Extracting the distance feature d and the angle situation feature T of the target maneuver from the target maneuver prediction data set S obtained in the step 1 a ：
In the formula, X u ,Y u ,Z u Respectively representing the three-dimensional spatial position coordinates, X, of my unmanned aerial vehicle e ,Y e ,Z e Respectively representing the three-dimensional spatial position coordinates of the target drone.
Merging the extracted distance features and angle situation features into a target maneuver prediction data set S:
S＝[s u ,c u ,s e ,c e ,d,T a ]
in the formula, s u ,c u Respectively representing the state information and maneuver control information, S, of the unmanned aerial vehicle of the same party in the target maneuver prediction data set S e ,c e Respectively representing the state information and the maneuver control information of the target unmanned aerial vehicle in the target maneuver prediction data set S.
The step 3 comprises the following steps:
3a, dividing the target maneuver prediction data set S in the step 1 into a training set S according to the proportion that the ratio of the training set to the test set is 7:3 train And test set S test ；
3b, respectively extracting a training set S train And test set S test All dimension data of the maneuvering control information are subjected to labeling pretreatment to obtain labeled training set control information data
And test set status informationInformation data
The unmanned aerial vehicle controls the speed and direction of the unmanned aerial vehicle through acceleration, pitch angle speed and yaw angle speed, the change of the speed of the unmanned aerial vehicle can be influenced by the magnitude of the acceleration, the existence of the pitch angle speed and the yaw angle speed controls the direction of the unmanned aerial vehicle, in order to simulate different maneuvers of the unmanned aerial vehicle in air battle, a simulation system sets 4 value conditions { -30, -10,40,60} (the unit is meter/second square) for the magnitude of the acceleration, each value condition represents that the unmanned aerial vehicle performs different acceleration or deceleration maneuvers, the pitch angle speed has 9 value conditions { -30, -21, -12, -5,0,5,12,21,30} (the unit is degree/second), each value represents that the unmanned aerial vehicle performs different maneuvers deviating from the original maneuvers at a certain angular speed, the yaw angle speed has 9 value conditions { -60, -34, -18, -8,0,8,18,34,60 (the unit is degree/second), each value represents that the unmanned aerial vehicle performs different maneuvers deviating from the original maneuvers at a certain angular speed, so that the maneuver parameters 4 × 9 values are the maneuvers of the maneuvers, each combination of the maneuvers represents a maneuver, and the maneuver is a prediction tag for the unmanned aerial vehicle, and the unmanned aerial vehicle control parameters are a combined maneuver, and the yaw angle speed is a prediction tag;
3c, respectively aligning the training sets S train And test set S test Carrying out normalization pretreatment on the data of each dimension to obtain normalized training set data
And test set data
In the formula, x i Data representing the ith dimension in a training set or a test set,
represents the minimum value of the ith-dimension data,
represents the maximum value of the ith-dimension data,
is the result of the normalization of the ith dimension data;
normalized training set data
And test set data
Are stored in a dimensional format of (batch, seq, input), wherein batch represents the dimension of the network model batch processing, seq represents the length of the time series used for predicting the maneuver of the target at the next moment, and input represents the dimension of each piece of data in the normalized data set.
The step 4 specifically comprises the following steps:
the maneuvering prediction network model based on the target of the LSTM comprises 1 LSTM network with 2 hidden layers, 3 full-connection layers and 3 cross entropy loss functions, wherein the input of the LSTM network model is a training set data sequence normalized by continuous seq time points, and the input format is (batch, seq, input); the output is the hidden state of the target machine maneuver in the packaged near seq time period
The output format is (batch, seq, hidden);
hidden state of target machine maneuver within approximate seq time period
Respectively inputting 3 full-connection layers, combining 3 cross entropy loss functions L a ，L p ，L y Respectively obtaining the probability condition that each type of value is taken at the moment of seq +1 by the acceleration, the pitch angle speed and the yaw angle speed:
in the formula, L represents a cross entropy function calculation formula, N represents the number of samples, M represents the number of categories, y ic Is a symbolic function, if the true class of the sample i is c, y ic Get 1, otherwise y ic Taking 0; l is a radical of an alcohol a ，L p ，L y Respectively representing acceleration cross entropy loss, pitch angle cross entropy loss and yaw angle cross entropy loss which are obtained by calculating acceleration, pitch angle speed and yaw angle speed by using a cross entropy loss function L; p a Representing the probability that the target acceleration takes each class of value at the time seq +1,
representing the probability that the target acceleration takes the mth class value at seq +1, P p Representing the probability that the target pitch angle velocity takes each type of value at the time seq +1,
representing the probability that the target pitch angle velocity takes the nth value at seq +1 y Representing the probability case that the target yaw rate takes each type of value at time seq +1,
the probability that the target yaw rate takes the nth-class value at seq +1 is shown.
Respectively taking the category values with the maximum value probability of the target acceleration a, the pitch angle speed P and the yaw angle speed y as the predicted category values of the target machine seq +1 moment acceleration a, the pitch angle speed P and the yaw angle speed y, P a pre ，P p pre ，P y pre ：
And combining the predicted category values of the acceleration a, the pitch angle speed p and the yaw angle speed y to obtain a predicted category label value of the maneuvering action, and converting the maneuvering prediction problem of the unmanned aerial vehicle into a classification prediction problem of the target acceleration a, the pitch angle speed p and the yaw angle speed y.
L train ＝w a L a +w p L p +w y L y
In the formula, w a ，w p ，w y Respectively representing the acceleration cross entropy loss L a Pitch angle cross entropy loss L p Sum yaw cross entropy loss L y The weight of (2).
An electronic device comprises a processor, a memory and a communication bus, wherein the processor and the memory are communicated with each other through the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the maneuvering prediction method based on the LSTM target when executing the program stored in the memory.
A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, implements a method of maneuver prediction for LSTM-based targets as described above.
The invention has the beneficial effects that:
when the maneuver of the target machine is predicted, the spatial situation occupation characteristics of the target machine are extracted, the spatial situation occupation characteristics, the state and the maneuver information of the unmanned aerial vehicle of the owner and the state and the maneuver information of the target machine are jointly used as the basis for the maneuver prediction of the target machine, the law of maneuver selection of the target machine in air combat countermeasure data can be fully excavated, and the maneuver prediction of the target machine is more accurate
The invention analyzes the relation between the maneuvering action of the unmanned aerial vehicle and the acceleration, pitch angle speed and deflection angle speed of the unmanned aerial vehicle, labels the values of the acceleration, pitch angle speed and deflection angle speed, converts the prediction problem of the maneuvering action selection of the target machine into the classification prediction problem of three maneuvering control variables of the acceleration, pitch angle speed and deflection angle speed, can accurately predict the future maneuvering action of the target machine, and provides a new idea for the research of the maneuvering prediction method of the target machine
The method starts from the characteristic that the target machine maneuver has time sequence, establishes the target machine maneuver prediction network model based on the LSTM, predicts the maneuver of the target machine at the next moment according to the state information and maneuver information of the unmanned aerial vehicle of one party in a period of time, the state information and maneuver information of the target machine and the space situation occupation characteristics of the target machine in the period of time by utilizing the strong prediction capability of the LSTM on time sequence data, obtains better prediction effect compared with other network models, and realizes the accurate prediction of the target machine maneuver.
Description of the drawings:
fig. 1 is a schematic view of the azimuth angles of the unmanned aerial vehicle and the target aircraft in the air combat countermeasure process.
Fig. 2 is a schematic diagram of labeling processing of unmanned aerial vehicle maneuvering control parameters.
FIG. 3 is a schematic diagram of a maneuver prediction method for an LSTM-based target.
Detailed Description
The present invention will be described in further detail with reference to examples.
The maneuvering prediction method of the target based on the LSTM comprises the following steps:
1, constructing a target aircraft maneuvering selection data set in the fight process of the unmanned aerial vehicle;
utilizing a certain one-to-one air combat simulation system, two combat parties set different initial states to carry out a plurality of combat experiments, and recording the unmanned aerial vehicle shapes of the two combat parties every 20ms in each combatState information s t And maneuver control information c t Air combat countermeasure data is formed and each combat process is terminated when the winning or the duration of the combat by a party exceeds 2 minutes. In this example, 280 groups of battle experiments are selected to include 14397712 pieces of battle maneuver data as the data set S of target maneuver prediction. At any time t, state information s of unmanned aerial vehicles of both parties of the battle t And maneuver control information c t Each contains the following information in multiple dimensions, as shown in the formula:
s t ＝[X t ,Y t ,Z t ,EX t ,EY t ,EZ t ,AX t ,AY t ,AZ t ,V t ]
c t ＝[a t ,p t ,y t ]
in the formula, X t ,Y t ,Z t Respectively representing the three-dimensional spatial position coordinates, EX, of the drone at time t t ,EY t ,EZ t Respectively representing the Euler angles of the postures of the unmanned aerial vehicle at the t moment (AX) t ,AY t ,AZ t ) The aircraft nose orientation vector at the moment t of the unmanned aerial vehicle, V t Speed information, s, representing the unmanned plane at time t t Containing 10 dimensions of state information. a is t ,p t ,y t Acceleration information, pitch angle information and yaw angle information, c, representing the speed and direction of the controlling drone, respectively t Contains motorized control information in 3 dimensions. State information of unmanned aerial vehicle at any moment t in the embodiment is used
For indicating, maneuvering control information
Representing; state information of target unmanned aerial vehicle
For indicating, for manoeuvre control information
Represents; data set S = [ S ] u ,c u ,s e ,c e ]。
2, extracting space situation characteristics of the target maneuver countermeasure process;
in the air combat countermeasure process, the unmanned aerial vehicle continuously maneuvers to carry out game countermeasure with the target unmanned aerial vehicle, so that the space occupation of the unmanned aerial vehicle is in an advantage state, and the aim of attacking the target unmanned aerial vehicle is fulfilled. Therefore, the spatial situation information is a key factor for the unmanned aerial vehicle to perform maneuver selection. In the short-distance air combat confrontation process, the key of the unmanned aerial vehicle winning is that when the target machine is in the attack range of the own unmanned aerial vehicle, the own unmanned aerial vehicle is in an advantageous state in angle situation, and the value of the azimuth angle p of the target machine and the value of the entrance angle q of the target machine at the moment are close to 0, as shown in fig. 1. Therefore, the distance d between the target machine and the angle situation T of the target machine need to be considered when the maneuver of the target machine is predicted a Extracting the distance characteristic d and the angle situation characteristic T of the target maneuver from the target maneuver prediction data set S a ：
In the formula, X u ,Y u ,Z u Respectively representing three-dimensional space position coordinates, X, of unmanned aerial vehicle of our party e ,Y e ,Z e Respectively representing the three-dimensional spatial position coordinates of the target drone.
Merging the extracted distance features and angle situation features into a target maneuver prediction data set S:
S＝[s u ,c u ,s e ,c e ,d,T a ]
in the formula, s u ,c u Respectively representing the state information and maneuver control information, S of the unmanned aerial vehicle at one party in the target maneuver prediction data set S e ,c e Respectively representing target maneuversAnd predicting the state information and the maneuvering control information of the target unmanned aerial vehicle in the data set S.
And 3, segmenting and preprocessing the data set.
3a, dividing the target maneuver prediction data set S into a training set S according to the ratio of the training set to the test set of 7:3 train And test set S test ；
3b, respectively extracting a training set S train And test set S test All dimension data of the maneuvering control information are subjected to labeling pretreatment to obtain labeled training set control information data
And test set status information data
Unmanned aerial vehicle passes through the speed and the direction of acceleration, pitch angle speed and yaw angular velocity control self, and the change of unmanned aerial vehicle speed can be influenced to the size of acceleration, and the direction of unmanned aerial vehicle is being controlled in the existence of pitch angle speed and yaw angular velocity. In order to simulate different maneuvers of the unmanned aerial vehicle in the air battle, the simulation system sets 4 value cases { -30, -10,40,60} (the unit is meter per second square), each value case represents that the unmanned aerial vehicle performs different acceleration or deceleration maneuvers, 9 value cases { -30, -21, -12, -5,0,5,12,21,30} (the unit is degree per second) are available in the pitching angular speed, each value case represents that the unmanned aerial vehicle performs different pitching maneuvers at a certain angular speed, 9 value cases { -60, -34, -18, -8,0,8,18,34,60} (the unit is degree per second) are available in the yaw angular speed, each value case represents that the unmanned aerial vehicle performs different deviation maneuvers from the original flight path at a certain angular speed, so that the maneuvering parameters of the unmanned aerial vehicle share 4 × 9=324 value combinations, each combination represents one maneuvering action of the unmanned aerial vehicle, and the prediction value of the maneuvering actions is the prediction of the maneuvering control parameter combinations. Values of the acceleration, the pitch angle velocity and the yaw angle velocity are mapped into a type label form as shown in fig. 2;
3c, respectively aligning the training sets S train And test set S test Carrying out normalization pretreatment on the data of each dimension to obtain normalized training set data
And test set data
In the formula, x i Data representing the ith dimension in a training set or a test set,
represents the minimum value of the ith-dimension data,
represents the maximum value of the ith-dimensional data,
is the result of the normalization of the ith dimension data;
normalized training set data
And test set data
Are stored in a dimensional format of (batch, seq, input), wherein batch represents the dimension of the network model batch processing, seq represents the length of the time series used for predicting the maneuver of the target at the next moment, and input represents the dimension of each piece of data in the normalized data set. In this example, batch is set to 256, seq is set to 14, input is 28.
4, establishing a maneuvering prediction network model of the target based on the LSTM;
as shown in FIG. 3, LSTM-based targetsThe maneuver prediction network model of (1) comprises 1 LSTM network with 2 hidden layers, 3 fully connected layers and 3 cross entropy loss functions. The input of the LSTM network model is a training set data sequence normalized at continuous seq moments, and the input format is (batch, seq, input); the output is the hidden state of the target machine maneuver in the packaged near seq time period
The output format is (batch, seq, hidden), hidden in this example is set to 128;
hidden state of target machine maneuver within near seq time period
Respectively inputting 3 full-connection layers, combining 3 cross entropy loss functions L a ，L p ，L y Respectively obtaining the probability condition of each type of value of the acceleration, the pitch angular velocity and the yaw angular velocity at seq + 1:
in the formula, L represents a cross entropy function calculation formula, N represents the number of samples, M represents the number of categories, y ic Is a symbolic function, if the true class of the sample i is c, y ic Get 1, otherwise y ic Taking 0; l is a ，L p ，L y Respectively representing acceleration cross entropy loss, pitch angle cross entropy loss and yaw angle cross entropy loss which are obtained by calculating acceleration, pitch angle speed and yaw angle speed by using a cross entropy loss function L; p a Representing the probability that the target acceleration takes each class of value at the time seq + 1,
representing the probability that the target acceleration takes the mth class value at the time of seq +1Situation, P p Representing the probability that the target pitch angle velocity takes each type of value at the time seq + 1,
representing the probability that the target pitch angle velocity takes the nth value at seq +1 y Representing the probability case that the target yaw rate takes each type of value at time seq + 1,
the probability that the target yaw rate takes the nth-class value at seq +1 is shown.
Respectively taking the category values with the maximum value probability of the target acceleration a, the pitch angle speed P and the yaw angle speed y as the predicted category values of the target machine seq +1 moment acceleration a, the pitch angle speed P and the yaw angle speed y, P a pre ，P p pre ，P y pre ：
And combining the predicted category values of the acceleration a, the pitch angle speed p and the yaw angle speed y to obtain a predicted category label value of the maneuvering action, and obtaining a corresponding maneuvering control variable value according to the mapping relation between the category label and the maneuvering control parameter value shown in fig. 2. To this end, the maneuver prediction problem of the unmanned aerial vehicle is converted into a classification prediction problem of the target acceleration a, the pitch angular velocity p and the yaw angular velocity y.
5, using the training set
Training a maneuver prediction network model for an LSTM-based target and utilizing a test set
Detecting the accuracy;
during model training, the cross entropy loss L of the acceleration is calculated a Pitch angle cross entropy loss L p Sum yaw angle cross entropy loss L y Performing weighted fusion, and taking the fused loss function value as the training loss L of the whole network model train ：
L train ＝w a L a +w p L p +w y L y
In the formula, w a ，w p ，w y Respectively representing the acceleration cross entropy loss L a Pitch angle cross entropy loss L p Sum yaw angle cross entropy loss L y In this example, the setting w is determined experimentally a ，w p ，w y Are 0.2,0.4,0.4, respectively.
In order to prevent the overfitting phenomenon in the model training process and make the network model more robust, a dropout mechanism is added in the LSTM network layer, and the value of dropout is set to be 0.4 in the example; in this example, the learning rate of the network model is set to 0.0001 and the number of training rounds epochs is set to 500. The network parameters of the model are saved for each round of training.
And testing and verifying the test set data by using the stored network model parameters, wherein the evaluation index adopted by the model is the Accuracy, and the more the Accuracy is close to 1, the more accurate the classification prediction of the model is. The invention respectively counts the prediction accuracy Acc (a), acc (p), acc (y) of the test centralized acceleration, the pitch angle speed and the yaw angle speed and the accuracy Acc (a, p, y) of the acceleration, the pitch angle speed and the yaw angle speed which are predicted correctly at the same time, simultaneously compares the model with a BP neural network and an RNN neural network, and performs independent experiments for each compared network model for many times, wherein the experiment results are shown in the table 1:
TABLE 1 comparison of different network models to target machine maneuver prediction accuracy
As can be seen from the results in table 1, in this example, compared with the BP neural network and the RNN, the network model of the present invention achieves good results for both the prediction of a single maneuver control parameter and the prediction of all maneuver control parameters at the same time, wherein the prediction of all maneuver control parameters at the same time indicates that the combined value of the maneuver control parameters is predicted correctly, i.e. the target maneuver at a certain time is predicted correctly.
The invention analyzes the target maneuver prediction method in the existing air combat countermeasure process, finds that the existing target maneuver prediction method only starts from the target, divides the maneuver of the target machine into the fixed weighted combination of several actions, and the weight value of each action represents the probability of the target machine taking the maneuver. The relation between the maneuvering action and the space situation occupied by the maneuvering action is not considered; the influence condition of the unmanned aerial vehicle state of the owner on the maneuvering of the target machine is not considered; therefore, when the maneuver of the target machine is predicted, the space situation occupation characteristics of the target machine are extracted, and the space situation occupation characteristics, the state and the maneuver information of the unmanned aerial vehicle of the same party and the state and the maneuver information of the target machine are used as the basis for the maneuver prediction of the target machine, so that the law of maneuver selection of the target machine in the air combat countermeasure data can be fully mined, and the maneuver prediction of the target machine is more accurate
The invention analyzes the relation between the maneuvering action of the unmanned aerial vehicle and the acceleration, pitch angle speed and deflection angle speed of the unmanned aerial vehicle, labels the values of the acceleration, pitch angle speed and deflection angle speed, converts the maneuvering prediction problem of the target machine into the classification prediction problem of three maneuvering control variables of the acceleration, pitch angle speed and deflection angle speed, can accurately predict the future maneuvering action of the target machine, and provides a new thought for the research of the maneuvering prediction method of the target machine
The invention sets up an LSTM-based target machine maneuver prediction network model based on the characteristic that the target machine maneuver has time sequence, and predicts the maneuver action of the target machine at the next moment according to the state information and maneuver information of the unmanned aerial vehicle of one party in a period of time, the state information and maneuver information of the target machine and the space situation occupation characteristic of the target machine in the period of time by utilizing the strong prediction capability of the LSTM on time sequence data. Due to space limitation, the prediction method shows a prediction example that in the process of resisting a certain air combat, a target machine actually conducts 256 maneuvers for a batch time at a certain time and correspondingly conducts 256 predicted maneuvers for the model, the prediction effect is shown in the table 2, and the experimental result further verifies the effectiveness of the prediction method for the maneuvering of the target machine.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on differences from other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (10)
1. A maneuver prediction method for an LSTM-based target, comprising the steps of;
step 1: constructing a target maneuvering selection data set S in the unmanned aerial vehicle fighting process, wherein the target is a target unmanned aerial vehicle in fighting with the unmanned aerial vehicle of one party;
step 2: extracting space situation characteristics of a target maneuver countermeasure process;
and step 3: segmenting and preprocessing the target mobile selection data set S in the step 1;
and 4, step 4: establishing a maneuvering prediction network model of the target based on the LSTM;
and 5: using training sets
Training a maneuver prediction network model for an LSTM-based target and utilizing a test set
And detecting the prediction accuracy.
2. The LSTM-based target maneuver prediction method of claim 1, wherein step 1 utilizes a one-to-one air combat simulation system, wherein the two parties set different initial states for carrying out a plurality of combat experiments, and the state information s of the unmanned aerial vehicles of the two parties is recorded every 20ms during each combat t And maneuver control information c t Forming air combat confrontation data, and stopping when the winning or the fighting duration of a certain party exceeds 2 minutes in each combat process;
selecting 280 groups of air combat countermeasure data as a data set S of target maneuver prediction, and at any time t, obtaining state information S of unmanned aerial vehicles of both parties of the battle t And maneuver control information c t Each contains the following information in multiple dimensions, as shown in the formula:
s t ＝[X t ,Y t ,Z t ,EX t ,EY t ,EZ t ,AX t ,AY t ,AZ t ,V t ]
c t ＝[a t ,p t ,y t ]
in the formula, X t ,Y t ,Z t Respectively representing the three-dimensional spatial position coordinates, EX, of the drone at time t t ,EY t ,EZ t Respectively representing the Euler angles of the postures of the unmanned aerial vehicle at the t moment (AX) t ,AY t ,AZ t ) Form the aircraft nose orientation vector at moment t of unmanned aerial vehicle, V t Speed information, s, representing the unmanned plane at time t t Containing status information of 10 dimensions, a t ,p t ,y t Acceleration information, pitch angle information and yaw angle information, c, representing the speed and direction of the controlling drone, respectively t Contains maneuvering control information of 3 dimensions, and is used for state information of unmanned aerial vehicle at any time t
For indicating, maneuvering control information
Represents; state information of target unmanned aerial vehicle
For indicating, maneuvering control information
Represents; data set
3. The method of claim 1 for maneuver prediction of LSTM-based targets, wherein the method further comprisesIn the step 2, in the process of the short-distance air combat countermeasure, the key of the unmanned aerial vehicle winning is that when the target machine is in the own unmanned aerial vehicle attack range, the own unmanned aerial vehicle angle situation is in the dominant state, the value of the azimuth angle p of the target machine and the value of the entrance angle q of the target machine at the moment are close to 0, and the distance d between the target machine and the own machine and the angle situation T of the target machine need to be considered when the maneuver of the target machine is predicted a Extracting the distance feature d and the angle situation feature T of the target maneuver from the target maneuver prediction data set S obtained in the step 1 a ：
In the formula, X u ,Y u ,Z u Respectively representing the three-dimensional spatial position coordinates, X, of my unmanned aerial vehicle e ,Y e ,Z e Respectively representing the three-dimensional spatial position coordinates of the target drone.
4. The method of claim 3, wherein the extracted distance features and angular situation features are incorporated into the target maneuver prediction dataset S:
S＝[s u ,c u ,s e ,c e ,d,T a ]
in the formula, s u ,c u Respectively representing the state information and maneuver control information, S, of the unmanned aerial vehicle of the same party in the target maneuver prediction data set S e ,c e And respectively representing the state information and the maneuver control information of the target unmanned aerial vehicle in the target maneuver prediction data set S.
5. The method of claim 1, wherein the step 3 comprises:
3a, dividing the target maneuver prediction data set S in the step 1 into a training set S according to the proportion that the ratio of the training set to the test set is 7:3 train And test set S test ；
3b, respectively extracting a training set S train And test set S test All dimension data of the maneuvering control information are subjected to labeling pretreatment to obtain labeled training set control information data
And test set status information data
The unmanned aerial vehicle controls the speed and the direction of the unmanned aerial vehicle through acceleration, pitch angle speed and yaw angle speed, the change of the speed of the unmanned aerial vehicle can be influenced by the magnitude of the acceleration, the direction of the unmanned aerial vehicle is controlled by the pitch angle speed and the yaw angle speed, and in order to simulate different maneuvers of the unmanned aerial vehicle in an air battle, the simulation system sets 4 value conditions { -30, -10,40,60} in the magnitude of the acceleration (the unit is: meter/second square), each value-taking value indicates that the unmanned aerial vehicle performs different acceleration or deceleration maneuvers, the pitch angle speed has 9 value-taking conditions { -30, -21, -12, -5,0,5,12,21,30} (the unit is: degree/second), each value-taking value indicates that the unmanned aerial vehicle performs different pitching maneuvers at a certain angular speed, the yaw angular speed has 9 value-taking conditions { -60, -34, -18, -8,0,8,18,34,60} (the unit is: degree/second), each value-taking value indicates that the unmanned aerial vehicle performs different maneuvers deviating from an original flight path at a certain angular speed, so that the maneuvering control parameters of the unmanned aerial vehicle have 4 value-taking combinations of 9 × 9=324, each combination represents one maneuvering action of the unmanned aerial vehicle, the maneuver action prediction of the maneuvering action is the prediction of the maneuvering control parameter combination value, and the values of the acceleration, the pitch angle speed and the yaw angular speed are mapped into a type label form;
3c, respectively aligning the training sets S train And test set S test Carrying out normalization pretreatment on the data of each dimension to obtain normalized training set data
And test set data
In the formula, x i Data representing the ith dimension in a training set or a test set,
represents the minimum value of the ith-dimension data,
represents the maximum value of the ith-dimension data,
is the result of the normalization of the ith dimension data;
normalized training set data
And test set data
Are stored in a dimensional format of (batch, seq, input), wherein batch represents the dimension of the network model batch processing, seq represents the length of the time series used for predicting the maneuver of the target at the next moment, and input represents the dimension of each piece of data in the normalized data set.
6. The method for maneuver prediction of LSTM-based targets according to claim 1, wherein said step 4 is specifically:
the model of the mobile prediction network of the LSTM-based target comprises 1 LSTM network with 2 hidden layers, 3 fully connected layers and3 cross entropy loss functions, wherein the input of the LSTM network model is a training set data sequence normalized at continuous seq time, and the input format is (batch, seq, input); the output is the hidden state of the target machine maneuver in the packaged near seq time period
The output format is (batch, seq, hidden);
hidden state of target machine maneuver within near seq time period
Respectively inputting 3 full-connection layers, combining 3 cross entropy loss functions L a ，L p ，L y Respectively obtaining the probability condition of each type of value of the acceleration, the pitch angular velocity and the yaw angular velocity at seq + 1:
in the formula, L represents a cross entropy function calculation formula, N represents the number of samples, M represents the number of categories, y ic Is a symbolic function, if the true class of the sample i is c, y ic Get 1, otherwise y ic Taking 0; l is a ，L p ，L y Respectively representing acceleration cross entropy loss, pitch angle cross entropy loss and yaw angle cross entropy loss which are obtained by calculating acceleration, pitch angle speed and yaw angle speed by using a cross entropy loss function L; p a Representing the probability that the target acceleration takes each class of value at the time seq +1,
representing the probability that the target acceleration takes the mth class value at seq +1, P p Representing target pitch angle velocity at seq +1 time takes the probability case of each class of value,
represents the probability that the target pitch angular velocity takes the nth value at seq +1 y Representing the probability case that the target yaw rate takes each type of value at time seq +1,
the probability that the target yaw rate takes the nth-class value at seq +1 is shown.
7. The LSTM-based target maneuvering prediction method as recited in claim 6, characterized in that the category values with the maximum value probability of the target acceleration a, the pitch angle velocity P and the yaw angle velocity y are respectively used as the predicted category values of the target machine seq +1 moment acceleration a, the pitch angle velocity P and the yaw angle velocity y, P a pre ，P p pre ，P y pre ：
And combining the predicted category values of the acceleration a, the pitch angle speed p and the yaw angle speed y to obtain a predicted category label value of the maneuvering action, and converting the maneuvering prediction problem of the unmanned aerial vehicle into a classification prediction problem of the target acceleration a, the pitch angle speed p and the yaw angle speed y.
8. The method of claim 1, wherein step 5 is implemented for cross-entropy loss L of acceleration during model training a Pitch angle cross entropy loss L p Sum yaw angle cross entropy loss L y Performing weighted fusion, and taking the fused loss function value as the training loss L of the whole network model train ：
L train ＝w a L a +w p L p +w y L y
In the formula, w a ，w p ，w y Respectively representing the acceleration cross entropy loss L a Pitch angle cross entropy loss L p Sum yaw angle cross entropy loss L y The weight of (2).
9. An electronic device is characterized by comprising a processor, a memory and a communication bus, wherein the processor and the memory are communicated with each other through the communication bus;
a memory for storing a computer program;
a processor, configured to execute a program stored in a memory, to implement a method of maneuver prediction based on the LSTM objective of any of claims 1 to 8.
10. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, implements a method of maneuver prediction for LSTM-based objectives according to any of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210986452.6A CN115480582A (en) | 2022-08-17 | 2022-08-17 | Maneuver prediction method for LSTM-based target, electronic device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210986452.6A CN115480582A (en) | 2022-08-17 | 2022-08-17 | Maneuver prediction method for LSTM-based target, electronic device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115480582A true CN115480582A (en) | 2022-12-16 |
Family
ID=84423028
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210986452.6A Pending CN115480582A (en) | 2022-08-17 | 2022-08-17 | Maneuver prediction method for LSTM-based target, electronic device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115480582A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116227952A (en) * | 2023-05-09 | 2023-06-06 | 中国人民解放军海军潜艇学院 | Method and device for selecting attack target defense strategy under key information deficiency |
| CN116383731A (en) * | 2023-03-06 | 2023-07-04 | 南京航空航天大学 | Tactical maneuver identification method, tactical maneuver identification system, electronic equipment and storage medium |
| CN117150738A (en) * | 2023-08-10 | 2023-12-01 | 中国船舶集团有限公司第七〇九研究所 | Action direction pre-judging method under complex scene |
-
2022
- 2022-08-17 CN CN202210986452.6A patent/CN115480582A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116383731A (en) * | 2023-03-06 | 2023-07-04 | 南京航空航天大学 | Tactical maneuver identification method, tactical maneuver identification system, electronic equipment and storage medium |
| CN116383731B (en) * | 2023-03-06 | 2023-11-14 | 南京航空航天大学 | Tactical maneuver identification method, tactical maneuver identification system, electronic equipment and storage medium |
| CN116227952A (en) * | 2023-05-09 | 2023-06-06 | 中国人民解放军海军潜艇学院 | Method and device for selecting attack target defense strategy under key information deficiency |
| CN116227952B (en) * | 2023-05-09 | 2023-07-25 | 中国人民解放军海军潜艇学院 | Method and device for selecting attack target defense strategy under key information deficiency |
| CN117150738A (en) * | 2023-08-10 | 2023-12-01 | 中国船舶集团有限公司第七〇九研究所 | Action direction pre-judging method under complex scene |
| CN117150738B (en) * | 2023-08-10 | 2024-05-10 | 中国船舶集团有限公司第七〇九研究所 | Action direction pre-judging method under complex scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115480582A (en) | Maneuver prediction method for LSTM-based target, electronic device and storage medium | |
| US11669110B2 (en) | Control system based on multi-unmanned aerial vehicle cooperative strategic confrontation | |
| CN109032168B (en) | DQN-based multi-unmanned aerial vehicle collaborative area monitoring airway planning method | |
| CN111399541B (en) | Unmanned aerial vehicle whole-region reconnaissance path planning method of unsupervised learning type neural network | |
| CN108731684B (en) | Multi-unmanned aerial vehicle cooperative area monitoring airway planning method | |
| EP2137481B1 (en) | A method and a system for estimating the impact area of a military load launched from an aircraft | |
| CN110782481A (en) | Unmanned ship intelligent decision method and system | |
| CN114048889A (en) | Aircraft trajectory prediction method based on long-term and short-term memory network | |
| CN110569792A (en) | Method for detecting front object of automatic driving automobile based on convolutional neural network | |
| Zhang et al. | Maneuver decision-making of deep learning for UCAV thorough azimuth angles | |
| CN114676743B (en) | Low-speed small target track threat identification method based on hidden Markov model | |
| CN117311393B (en) | Unmanned aerial vehicle autonomous flight path planning method and system | |
| CN116907282B (en) | Unmanned target aircraft ultra-low altitude flight control method based on artificial intelligence algorithm | |
| Xie et al. | Dual-channel and bidirectional neural network for hypersonic glide vehicle trajectory prediction | |
| CN112749761A (en) | Enemy combat intention identification method and system based on attention mechanism and recurrent neural network | |
| CN116432514A (en) | Interception intention recognition strategy simulation system and method for unmanned aerial vehicle attack and defense game | |
| Han et al. | DP-Authentication: A novel deep learning based drone pilot authentication scheme | |
| CN111401190A (en) | Vehicle detection method, device, computer equipment and storage medium | |
| CN114721423B (en) | Distribution method for multi-unmanned aerial vehicle to cooperatively reach preset target in consideration of collision avoidance constraint | |
| CN115661576A (en) | Method for identifying airplane group intention under sample imbalance | |
| CN115311580A (en) | Unmanned aerial vehicle threat determination method based on image recognition and related equipment | |
| CN115659229B (en) | Low, small and slow target threat degree evaluation method and device | |
| Sheng et al. | Target Threat Assessment in Air Combat with BP Neural Network for UAV | |
| CN109658742A (en) | The autonomous conflict Resolution method of intensively flight based on preamble flight information | |
| CN112327918B (en) | Multi-swarm sea area environment self-adaptive search algorithm based on elite learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |