CN116880198B - Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle - Google Patents
Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle Download PDFInfo
- Publication number
- CN116880198B CN116880198B CN202310917386.1A CN202310917386A CN116880198B CN 116880198 B CN116880198 B CN 116880198B CN 202310917386 A CN202310917386 A CN 202310917386A CN 116880198 B CN116880198 B CN 116880198B
- Authority
- CN
- China
- Prior art keywords
- feature vector
- time sequence
- flying
- speed
- vector
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 25
- 239000013598 vector Substances 0.000 claims description 238
- 230000003993 interaction Effects 0.000 claims description 70
- 230000003044 adaptive effect Effects 0.000 claims description 36
- 238000012937 correction Methods 0.000 claims description 26
- 238000013527 convolutional neural network Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 18
- 238000009826 distribution Methods 0.000 claims description 15
- 230000004927 fusion Effects 0.000 claims description 15
- 230000014759 maintenance of location Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 5
- 238000005457 optimization Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 description 17
- 238000010586 diagram Methods 0.000 description 14
- 230000008859 change Effects 0.000 description 8
- 238000003062 neural network model Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 238000012886 linear function Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
A power equipment self-adaptive control system for supersonic unmanned aerial vehicle and its method are disclosed. The system comprises: the flight parameter acquisition module is used for acquiring flight height values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and the engine thrust matching module is used for adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height value and the flying speed value of the supersonic unmanned aerial vehicle at a plurality of preset time points. In this way, the thrust output of a suitable aircraft engine can be adaptively matched based on the altitude and speed of flight, so that the supersonic unmanned aerial vehicle can have a thrust supply meeting mission requirements at different altitudes and different speeds of flight, meeting its thrust requirements under various flight conditions.
Description
Technical Field
The present disclosure relates to the field of intelligent control, and more particularly, to a power equipment adaptive control system for a supersonic unmanned aerial vehicle and a method thereof.
Background
A supersonic unmanned aerial vehicle is one that is capable of flying beyond sonic speeds. Supersonic flight means that its flight speed exceeds the speed of sound (about 343 meters per second or 1225 kilometers per hour), and unmanned vehicles of this type typically have a high degree of mobility and speed, enabling them to reach a destination quickly in a short time. Supersonic unmanned aerial vehicles are commonly used for military reconnaissance, surveillance, and batting tasks. They can cover a larger area in a shorter time and acquire intelligence or perform a striking action at a high speed. Because of its high speed and maneuverability, the supersonic unmanned plane can quickly traverse the enemy defense system, providing a battlefield advantage.
The supersonic unmanned plane needs to have enough thrust to overcome the resistance in the flight process, and the engine suction flow is different due to different gas densities at different heights in the flight task execution process, so that the generated thrust is different, and the flight resistance of the plane is different at different flight speeds. Therefore, the engine of the supersonic unmanned aerial vehicle is required to meet the thrust requirements of the supersonic unmanned aerial vehicle under various flight working conditions according to the flight task requirements.
Accordingly, a power plant adaptive control system for a supersonic unmanned aerial vehicle is desired.
Disclosure of Invention
In view of this, the present disclosure proposes a power equipment adaptive control system for a supersonic unmanned aerial vehicle and a method thereof, which can adaptively match the thrust output of an appropriate aircraft engine based on the flying altitude and the flying speed, so that the supersonic unmanned aerial vehicle can have a thrust supply meeting the mission requirement at different flying speeds and different flying speeds, and meet the thrust requirement thereof under various flying conditions.
According to an aspect of the present disclosure, there is provided a power equipment adaptive control system for a supersonic unmanned aerial vehicle, comprising:
the flight parameter acquisition module is used for acquiring flight height values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and
and the engine thrust matching module is used for adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height values and the flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points.
According to another aspect of the present disclosure, there is provided a power equipment adaptive control method for a supersonic unmanned aerial vehicle, including:
acquiring flying height values and flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and
and based on the flying height values and flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points, the thrust output value of the supersonic unmanned aerial vehicle is adaptively adjusted.
According to an embodiment of the present disclosure, the system includes: the flight parameter acquisition module is used for acquiring flight height values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and the engine thrust matching module is used for adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height value and the flying speed value of the supersonic unmanned aerial vehicle at a plurality of preset time points. In this way, the thrust output of a suitable aircraft engine can be adaptively matched based on the altitude and speed of flight, so that the supersonic unmanned aerial vehicle can have a thrust supply meeting mission requirements at different altitudes and different speeds of flight, meeting its thrust requirements under various flight conditions.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 shows a block diagram of a power plant adaptive control system for a supersonic drone, according to an embodiment of the present disclosure.
Fig. 2 shows a block diagram of the engine thrust matching module in a power-plant adaptive control system for a supersonic drone, according to an embodiment of the present disclosure.
Fig. 3 shows a block diagram of the feature extraction unit in a power equipment adaptive control system for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Fig. 4 shows a block diagram of the thrust output value control unit in the power equipment adaptive control system for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Fig. 5 shows a block diagram of the feature distribution corrector subunit in a power-plant adaptive control system for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Fig. 6 shows a flowchart of a power equipment adaptive control method for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Fig. 7 shows an architectural schematic diagram of a power equipment adaptive control method for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Fig. 8 illustrates an application scenario diagram of a power equipment adaptive control system for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this disclosure and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following detailed description in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order not to obscure the present disclosure.
As described above, the supersonic unmanned aerial vehicle needs to have enough thrust to overcome the resistance in the flight process, the engine suction flow is different due to different gas densities at different heights in the process of executing the flight task, the generated thrust is different, and the flight resistance of the aircraft is different at different flight speeds. Therefore, the engine of the supersonic unmanned aerial vehicle is required to meet the thrust requirements of the supersonic unmanned aerial vehicle under various flight working conditions according to the flight task requirements.
In view of the above technical needs, the technical concept of the present application is to adaptively match the thrust output of an appropriate aircraft engine based on the flying altitude and the flying speed, so that the supersonic unmanned aerial vehicle can have the thrust supply meeting the mission requirements at different flying speeds and different flying speeds, and meet the thrust requirements under various flying working conditions.
Based on the technical conception, the technical scheme of the method is that the flying height value and the flying speed value of the controlled supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period are firstly obtained. It should be appreciated that supersonic unmanned aerial vehicles need to have sufficient thrust to overcome drag during flight, and that at different altitudes during execution of flight tasks, due to different gas densities, the engine intake flow will be different, and thus the thrust produced will be different, and the aircraft will have different flight drag at different flight speeds. Based on the above, in the technical scheme of the application, firstly, a time sequence of flying height values and a time sequence of flying speed values of the supersonic unmanned aerial vehicle are collected.
Accordingly, fig. 1 shows a block diagram schematic of a power plant adaptive control system for a supersonic drone according to an embodiment of the present disclosure. As shown in fig. 1, a power equipment adaptive control system 100 for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure includes: the flight parameter acquisition module 110 is configured to acquire flight altitude values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of predetermined time points within a predetermined time period; and an engine thrust matching module 120, configured to adaptively adjust a thrust output value of the supersonic unmanned aerial vehicle based on the flying height values and the flying speed values of the supersonic unmanned aerial vehicle at the plurality of predetermined time points.
And then, arranging the flying height values and flying speed values of the controlled supersonic unmanned aerial vehicle at a plurality of preset time points into flying height time sequence input vectors and flying speed time sequence input vectors according to time dimensions, and extracting flying height time sequence feature vectors and flying speed time sequence feature vectors from the flying height time sequence input vectors and the flying speed time sequence input vectors. That is, the time sequence of the flying height value and the time sequence of the flying speed value of the supersonic unmanned aerial vehicle are subjected to data structuring according to a time dimension, and a flying height time sequence feature vector and a flying speed time sequence feature vector are extracted from the flying height time sequence input vector and the flying speed time sequence input vector, wherein the flying height time sequence feature vector is used for representing a time sequence change rule of the flying height (reflecting the time sequence change rule of the suction flow of an engine), and the flying speed time sequence feature vector is used for representing a time sequence change rule of the flying speed (reflecting the time sequence change rule of the flying resistance).
More specifically, in an embodiment of the present disclosure, as shown in fig. 2, the engine thrust matching module 120 includes: an input vector arrangement unit 121, configured to arrange the flying height values and flying speed values of the controlled supersonic unmanned aerial vehicle at the plurality of predetermined time points into flying height time sequence input vectors and flying speed time sequence input vectors according to a time dimension, respectively; a feature extraction unit 122 for extracting a flight altitude time series feature vector and a flight speed time series feature vector from the flight altitude time series input vector and the flight speed time series input vector; a feature interaction unit 123, configured to perform feature interaction on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector; a thrust output value control unit 124 for generating a recommended thrust output value of the aircraft engine based on the flying height-flying speed interaction feature vector; and an adaptive adjustment unit 125, configured to adaptively adjust a thrust output value of the supersonic unmanned aerial vehicle based on the recommended thrust output value of the aircraft engine.
In one specific example of the present application, extracting the altitude-time-sequence feature vector and the flight-speed-time-sequence feature vector from the altitude-time-sequence input vector and the flight-speed-time-sequence input vector includes: the flying height time sequence input vector passes through a flying height time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying height time sequence feature vector; and passing the flying speed time sequence input vector through a flying speed time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying speed time sequence feature vector. That is, using a flying height time sequence feature extractor and a flying speed time sequence feature extractor based on a one-dimensional convolution neural network model to perform one-dimensional convolution encoding on the flying height time sequence input vector and the flying speed time sequence input vector so as to capture the association features of flying height and flying speed in a local time sequence neighborhood in the flying height time sequence input vector and the flying speed time sequence input vector to obtain the flying height time sequence feature vector and the flying speed time sequence feature vector. Accordingly, in one possible implementation manner, as shown in fig. 3, the feature extraction unit 122 includes: a height feature extraction subunit 1221, configured to pass the height time sequence input vector through a height time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the height time sequence feature vector; and a speed feature extraction subunit 1222 for passing the flying speed time sequence input vector through a flying speed time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying speed time sequence feature vector.
Therefore, the feature extraction unit is used for performing one-dimensional convolution coding on the flying height time sequence input vector and the flying speed time sequence input vector by using the flying height time sequence feature extractor and the flying speed time sequence feature extractor based on the one-dimensional convolution neural network model, so as to extract the flying height time sequence feature vector and the flying speed time sequence feature vector. The two subunits of the feature extraction unit are a height feature extraction subunit and a speed feature extraction subunit, respectively. And the altitude characteristic extraction subunit performs one-dimensional convolution coding on the altitude time sequence input vector through an altitude time sequence characteristic extractor based on a one-dimensional convolution neural network model to obtain an altitude time sequence characteristic vector. The speed characteristic extraction subunit also carries out one-dimensional convolution coding on the flying speed time sequence input vector through a flying speed time sequence characteristic extractor based on the one-dimensional convolution neural network model to obtain a flying speed time sequence characteristic vector. The feature extraction unit is used for capturing relevant features in a local time sequence neighborhood in the time sequence input vector of the flying height and the flying speed through the convolutional neural network model. The convolutional neural network has good feature extraction capability in time sequence data, and features on different time scales can be captured through convolution operation. By inputting the flying height and flying speed time sequence input vectors into the corresponding feature extractors, flying height time sequence feature vectors and flying speed time sequence feature vectors with high expressive power and discrimination can be extracted. These feature vectors may be used for subsequent tasks or analysis, such as flight status identification, anomaly detection, prediction, and the like. The function of the feature extraction unit is to provide a more representative and useful representation of the features for subsequent tasks, thereby improving the performance and effectiveness of the model.
It should be noted that the one-dimensional convolutional neural network model is a neural network model for processing one-dimensional sequence data. Compared with the traditional fully-connected neural network, the one-dimensional convolutional neural network can better capture the local mode and the time sequence relation in the input sequence. In a one-dimensional convolutional neural network, the input data is represented as one-dimensional vectors, such as a flight altitude time sequence input vector and a flight speed time sequence input vector. One-dimensional convolutional neural networks utilize convolutional layers to extract features in the input data. The convolution layer scans over the input sequence by sliding a small window (convolution kernel) and computes the convolution operation of the data within the window with the convolution kernel. This captures the local pattern in the input sequence. One-dimensional convolutional neural networks also include other types of layers, such as pooling layers and fully-connected layers. The pooling layer serves to reduce the size of the feature map and may extract more important features. The fully connected layer is used to map the output of the last convolutional layer to the desired output dimension. In the flying height time sequence feature extractor and the flying speed time sequence feature extractor, a one-dimensional convolution neural network model is used for carrying out one-dimensional convolution coding on flying height time sequence input vectors and flying speed time sequence input vectors so as to capture the associated features of flying height and flying speed in the local time sequence neighborhood. Thus, the flight altitude time sequence characteristic vector and the flight speed time sequence characteristic vector can be obtained and used for subsequent tasks or analysis. It should be appreciated that the one-dimensional convolutional neural network model has the following roles in the sequence data processing: through convolution operation, the one-dimensional convolution neural network can extract features with different sizes from an input sequence, and the features can capture local modes and time sequence relations in sequence data, so that the input sequence can be better understood and represented; the dimension of the input sequence can be reduced by the one-dimensional convolution neural network through convolution and pooling operation, so that the parameter number and the calculation complexity of the model are reduced, and the efficiency and the reasoning speed of the model are improved; the prediction and classification, the one-dimensional convolutional neural network can be used for the task of predicting and classifying the sequence data, and the model can predict future sequences or classify the sequences into different categories by learning the characteristics and modes in the input sequences; sequence modeling, a one-dimensional convolutional neural network can be used for modeling and generating sequence data, and a model can generate new sequence data by learning the distribution and rules of an input sequence.
And further, performing feature interaction on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector. That is, the flying height time sequence feature vector (time sequence change rule of the engine suction flow) and the flying speed time sequence feature vector (time sequence change rule reflecting the flying resistance) are fused in a feature interaction mode to obtain the flying height-flying speed interaction feature vector. In one specific example of the present application, a cascading function is used to perform feature interactions between the altitude-time-sequence feature vector and the flight-speed-time-sequence feature vector to obtain an altitude-flight-speed interaction feature vector.
It should be appreciated that using a cascading function to perform a feature interaction between the fly-height timing feature vector and the fly-speed timing feature vector has several roles: the cascade function can connect the time sequence feature vector of the flying height and the time sequence feature vector of the flying speed together in sequence by fusing the related information, so that the relevance of the flying height and the flying speed in the flying process can be more comprehensively described, and better understanding of flying performance and flying characteristics is facilitated; providing more input features, performing feature interaction through cascading functions, and increasing the dimension of the input features, so that more information and richer feature representation can be provided for subsequent tasks or analysis, and the performance and expression capability of the model can be improved; enhancing feature relevance: the cascading function can directly connect the flying height time sequence feature vector and the flying speed time sequence feature vector together, so that the flying height time sequence feature vector and the flying speed time sequence feature vector are more closely related in feature space, which is helpful for a model to better capture time sequence relation and interaction between flying height and flying speed, and modeling capacity of flying performance is improved. Feature interactions between the altitude time sequence feature vector and the flight speed time sequence feature vector using cascading functions can provide more comprehensive and rich feature representations, enhance feature relevance, and provide more information and input features for subsequent tasks or analysis, which helps to improve the performance of the model and the modeling capability of the flight performance.
Accordingly, in one possible implementation manner, the feature interaction unit 123 is configured to: performing feature interaction between the flying height time sequence feature vector and the flying speed time sequence feature vector by using a cascading function to obtain a flying height-flying speed interaction feature vector; wherein the cascade function is formulated as:wherein (1)>And->All representing a point convolution of the input,/->To activate the function +.>Representing a splicing operation->Characteristic values representing respective positions in the flying height timing characteristic vector, +.>And the characteristic value of each position in the time sequence characteristic vector of the flying speed is represented.
Accordingly, in one possible implementation, as shown in fig. 4, the thrust output value control unit 124 includes: the fusion subunit 1241 is configured to perform forward propagation information retention fusion on the flight altitude time sequence feature vector and the flight speed time sequence feature vector to obtain a correction feature vector; a feature distribution corrector unit 1242, configured to perform feature distribution correction on the flying height-flying speed interaction feature vector based on the correction feature vector, so as to obtain a corrected flying height-flying speed interaction feature vector; and a decoding regression subunit 1243, configured to perform decoding regression on the corrected altitude-flight speed interaction feature vector through a decoder to obtain a decoded value, where the decoded value is used to represent the recommended thrust output value of the aircraft engine.
In particular, in the technical solution of the present application, the flying height time sequence feature vector and the flying speed time sequence feature vector respectively express one-dimensional local time sequence correlation features of the flying height value and the flying speed value, and due to the difference of source data, the distributions of the flying height value and the flying speed value in time sequence are inconsistent with each other, and in consideration of different influences of numerical noise, significant feature distribution misalignment exists between the flying height time sequence feature vector and the flying speed time sequence feature vector.
In this way, when the cascade function is used to perform the feature interaction between the flying height time sequence feature vector and the flying speed time sequence feature vector, the misalignment of the distribution between the valve opening value correlation feature vector and the feature of the valve change rate correlation feature vector between the valves can generate the loss of the feature information respectively expressed during the forward propagation of the model due to the point convolution and the activation operation during the feature interaction, thereby affecting the expression effect of the flying height-flying speed interaction feature vector on the time sequence correlation feature of the flying height value and the flying speed value, and also affecting the accuracy of the classification result obtained by the classifier of the flying height-flying speed interaction feature vector.
Based on this, the applicant of the present application determined by the time sequence feature vector of the flying heightAnd the flight speed timing feature vector +.>Performing forward propagation information preserving fusion to obtain corrected feature vectors, e.g. denoted as/>。
Accordingly, in one possible implementation, the fusion subunit 1241 is configured to: carrying out forward propagation information retention fusion on the flying height time sequence feature vector and the flying speed time sequence feature vector by using the following fusion optimization formula to obtain the correction feature vector; the fusion optimization formula is as follows:wherein (1)>Is the flight altitude time sequence feature vector, < >>Is the time sequence feature vector of the flying speed, +.>And->Respectively represent the left shift of the feature vector +.>Bit and right shift->Bit (s)/(s)>For rounding function, ++>Is the average of all eigenvalues of the altitude time sequence eigenvector and the flight speed time sequence eigenvector, +.>Representing a norm of the feature vector, +.>Is the distance between the altitude-time-sequence feature vector and the flight-speed-time-sequence feature vector, and +.>As a logarithmic function with base 2 +.>Andrespectively representing subtraction and addition by position, +.>And->For weighting superparameters, < >>Is the correction feature vector.
Here, for the flying height timing feature vectorAnd the flight speed timing feature vector +.>In the forward propagation process in the network model, floating point distribution errors and information losses on vector scale due to feature interaction operations are balanced and standardized by introducing a bitwise displacement operation of vectors from the viewpoint of uniformizing information, and distribution diversity is introduced by reshaping the distribution of feature parameters before feature interaction, thereby information retention (retrieval) is performed in a manner of expanding information entropy. Thus, the correction feature vector +.>The linear interpolation is carried out, and then the dot multiplication is carried out on the flying height-flying speed interaction feature vector, so that the number of the flying height-flying speed interaction feature vectors can be reducedAnd the information loss of the flight altitude-flight speed interaction feature vector on the expression of the time sequence correlation features of the flight altitude value and the flight speed value is improved, so that the accuracy of a classification result obtained by the classifier is improved.
Accordingly, in one possible implementation, as shown in fig. 5, the feature distribution corrector subunit 1242 includes: a linear interpolation secondary subunit 12421, configured to perform linear interpolation on the correction feature vector so that the correction feature vector and the altitude-flying speed interaction feature vector have the same scale; and a per-position point multiplication secondary subunit 12422, configured to calculate a per-position point multiplication between the correction feature vector and the altitude-speed interaction feature vector to obtain the correction altitude-speed interaction feature vector.
It should be appreciated that linear interpolation is a method for estimating an unknown data point between two known data points, based on a linear relationship assumption that the change in data between two known data points is linear, by using a linear function between the known data points to estimate the value of the unknown data point. Specifically, given two known data points (x 1, y 1) and (x 2, y 2), where x1 and x2 are known input values and y1 and y2 are corresponding output values, linear interpolation can calculate the value of the unknown data point by the following formula:where x is the input value of the unknown data point and y is the estimated output value. This formula effectively uses the slope between known data points to estimate the output value of the unknown data point.
It should be understood that the multiplication by position point refers to the operation of multiplying elements at corresponding positions of two vectors to obtain a new vector, and the multiplication by position point may also be regarded as multiplication by corresponding elements of a matrix. The realization of the point-by-position multiplication only needs to multiply the elements of the corresponding positions of the two vectors. Assuming that there are two vectors a and B of equal length, the resultant vector C multiplied by position point can be expressed as:wherein A [ i ]]Represents the ith element of vector A, B [ i ]]Represents the i-th element of the vector B, and n represents the length of the vector. There are many applications of the per-position point multiplication in the feature processing, in the above example, the per-position point multiplication is used to calculate the per-position point multiplication between the correction feature vector and the flying height-flying speed interaction feature vector, and the correction flying height-flying speed interaction feature vector is obtained, which can emphasize the importance of the two feature vectors at the same position, thereby extracting the feature with more distinction. By multiplying by the position points, local relations between feature vectors, such as correlations between features at adjacent times in the time series data, can be captured. The operation is often used in the tasks of feature fusion, feature interaction and the like in the convolutional neural network and other models, and is helpful for extracting richer feature representations.
Further, a recommended thrust output value of the aircraft engine is generated based on the altitude-speed interaction feature vector. In a specific example of the present application, the altitude-speed interaction feature vector is decoded back by a decoder to obtain a decoded value, where the decoded value is used to represent a recommended thrust output of the aircraft engine. That is, a non-linear function mapping model between the flight power and the flight conditions of the supersonic unmanned aerial vehicle is constructed by a decoder to decode and regress the altitude-speed interaction feature vector to obtain a decoded value representing the thrust output of the recommended aircraft engine. Finally, based on the recommended thrust output value of the aircraft engine, the thrust output value of the supersonic unmanned aerial vehicle is adaptively adjusted.
Accordingly, in one possible implementation, the decoding regression subunit 1243 is configured to: performing decoding regression on the corrected fly-height-fly-speed interaction feature vector in the following decoding formula using a plurality of full connection layers of the decoder to obtain the decoded value; wherein, the decoding formula is:wherein (1)>Is said corrected fly-height-fly-speed interaction feature vector,/>Is the decoded value,/->Is a weight matrix, < >>Is a bias vector, ++>Representing matrix multiplication->To activate the function.
In this way, the thrust output of the aircraft engine is adaptively matched based on the flight altitude and the flight speed, so that the supersonic unmanned aerial vehicle can have the thrust supply meeting the task requirements at different high speeds and different flight speeds, and the thrust requirements of the supersonic unmanned aerial vehicle under various flight working conditions are met.
In summary, a power-plant adaptive control system 100 for a supersonic unmanned aerial vehicle is illustrated that may adaptively match the thrust output of an appropriate aircraft engine based on altitude and speed of flight, such that the supersonic unmanned aerial vehicle may have a thrust supply that meets mission requirements at different altitudes and different speeds of flight, meeting its thrust requirements at various flight conditions.
As described above, the power equipment adaptive control system 100 for a supersonic unmanned aerial vehicle according to the embodiment of the present disclosure may be implemented in various terminal devices, for example, a server having a power equipment adaptive control algorithm for a supersonic unmanned aerial vehicle, or the like. In one example, the power equipment adaptive control system 100 for a supersonic drone may be integrated into the terminal device as one software module and/or hardware module. For example, the power equipment adaptive control system 100 for a supersonic unmanned aerial vehicle may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the power equipment adaptive control system 100 for a supersonic unmanned aerial vehicle may also be one of a plurality of hardware modules of the terminal device.
Alternatively, in another example, the power equipment adaptive control system for a supersonic unmanned aerial vehicle 100 and the terminal device may be separate devices, and the power equipment adaptive control system for a supersonic unmanned aerial vehicle 100 may be connected to the terminal device through a wired and/or wireless network and transmit the interactive information according to an agreed data format.
Fig. 6 shows a flowchart of a power equipment adaptive control method for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure. Fig. 7 shows a schematic diagram of a system architecture of a power equipment adaptive control method for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure. As shown in fig. 6 and 7, a power equipment adaptive control method for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure includes: s110, acquiring flight height values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and S120, adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height values and flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points.
In one possible implementation, adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height value and the flying speed value of the supersonic unmanned aerial vehicle at the plurality of predetermined time points includes: respectively arranging the flying height values and flying speed values of the controlled supersonic unmanned aerial vehicle at a plurality of preset time points into flying height time sequence input vectors and flying speed time sequence input vectors according to time dimensions; extracting a flight altitude time sequence feature vector and a flight speed time sequence feature vector from the flight altitude time sequence input vector and the flight speed time sequence input vector; performing feature interaction on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector; generating a recommended thrust output value of the aircraft engine based on the altitude-speed interaction feature vector; and adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the recommended thrust output value of the aircraft engine.
Here, it will be understood by those skilled in the art that the specific operations of the respective steps in the above-described power equipment adaptive control method for a supersonic unmanned aerial vehicle have been described in detail in the above description of the power equipment adaptive control system for a supersonic unmanned aerial vehicle with reference to fig. 1 to 5, and thus, repetitive descriptions thereof will be omitted.
Fig. 8 illustrates an application scenario diagram of a power equipment adaptive control system for a supersonic unmanned aerial vehicle according to an embodiment of the present disclosure. As shown in fig. 8, in this application scenario, first, the flying height values (e.g., D1 illustrated in fig. 8) and flying speed values (e.g., D2 illustrated in fig. 8) of the supersonic unmanned aerial vehicle at a plurality of predetermined time points within a predetermined period of time are acquired, and then the flying height values and flying speed values of the supersonic unmanned aerial vehicle at the plurality of predetermined time points are input into a server (e.g., S illustrated in fig. 8) in which a power equipment adaptive control algorithm for the supersonic unmanned aerial vehicle is deployed, wherein the server is capable of processing the flying height values and flying speed values of the supersonic unmanned aerial vehicle at the plurality of predetermined time points using the power equipment adaptive control algorithm for the supersonic unmanned aerial vehicle to obtain decoded values for representing the thrust output values of the recommended aircraft engine.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (5)
1. A power equipment adaptive control system for a supersonic unmanned aerial vehicle, comprising:
the flight parameter acquisition module is used for acquiring flight height values and flight speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and
the engine thrust matching module is used for adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the flying height values and the flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points;
wherein, engine thrust matching module includes:
the input vector arrangement unit is used for respectively arranging the flying height values and the flying speed values of the controlled supersonic unmanned aerial vehicle at a plurality of preset time points into flying height time sequence input vectors and flying speed time sequence input vectors according to the time dimension;
a feature extraction unit for extracting a flight altitude time sequence feature vector and a flight speed time sequence feature vector from the flight altitude time sequence input vector and the flight speed time sequence input vector;
the feature interaction unit is used for carrying out feature interaction on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector;
a thrust output value control unit for generating a recommended thrust output value of the aircraft engine based on the flying height-flying speed interaction feature vector; and
the self-adaptive adjusting unit is used for self-adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle based on the recommended thrust output value of the aircraft engine;
wherein the feature extraction unit includes:
the altitude characteristic extraction subunit is used for enabling the flight altitude time sequence input vector to pass through an altitude time sequence characteristic extractor based on a one-dimensional convolutional neural network model to obtain the flight altitude time sequence characteristic vector; and
a speed feature extraction subunit, configured to pass the flying speed time sequence input vector through a flying speed time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying speed time sequence feature vector;
wherein, the feature interaction unit is used for:
performing feature interaction between the flying height time sequence feature vector and the flying speed time sequence feature vector by using a cascading function to obtain a flying height-flying speed interaction feature vector;
wherein the cascade function is formulated as:wherein,and->All representing a point convolution of the input,/->To activate the function +.>The operation of the splice is indicated and,characteristic values representing respective positions in the flying height timing characteristic vector, +.>Characteristic values representing the positions in the time sequence characteristic vector of the flying speed;
wherein the thrust output value control unit includes:
the fusion subunit is used for carrying out forward propagation information retention fusion on the flying height time sequence feature vector and the flying speed time sequence feature vector so as to obtain a correction feature vector;
the characteristic distribution correction subunit is used for carrying out characteristic distribution correction on the flying height-flying speed interaction characteristic vector based on the correction characteristic vector so as to obtain a corrected flying height-flying speed interaction characteristic vector; and
and the decoding regression subunit is used for carrying out decoding regression on the corrected flying height-flying speed interaction characteristic vector through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended thrust output value of the aircraft engine.
2. The power equipment adaptive control system for a supersonic unmanned aerial vehicle of claim 1, wherein the fusion subunit is configured to:
carrying out forward propagation information retention fusion on the flying height time sequence feature vector and the flying speed time sequence feature vector by using the following fusion optimization formula to obtain the correction feature vector;
the fusion optimization formula is as follows:wherein,is the flight altitude time sequence feature vector, < >>Is the time sequence feature vector of the flying speed, +.>And->Respectively represent the left shift of the feature vector +.>Bit and right shift->Bit (s)/(s)>For rounding function, ++>Is the average of all eigenvalues of the altitude time sequence eigenvector and the flight speed time sequence eigenvector, +.>Representing a norm of the feature vector, +.>Is the distance between the altitude-time-sequence feature vector and the flight-speed-time-sequence feature vector, and +.>As a logarithmic function with base 2 +.>And->Respectively representing subtraction and addition by position, +.>And->For weighting superparameters, < >>Is the correction feature vector.
3. The power plant adaptive control system for a supersonic unmanned aerial vehicle according to claim 2, wherein the characteristic distribution corrector subunit comprises:
a linear interpolation secondary subunit, configured to perform linear interpolation on the correction feature vector so that the correction feature vector and the flying height-flying speed interaction feature vector have the same scale; and
and the secondary sub-unit is multiplied by the position point and used for calculating the position point multiplication between the correction feature vector and the flying height-flying speed interaction feature vector to obtain the correction flying height-flying speed interaction feature vector.
4. The power plant adaptive control system for a supersonic unmanned aerial vehicle of claim 3, wherein the decoding regression subunit is configured to:
performing decoding regression on the corrected fly-height-fly-speed interaction feature vector in the following decoding formula using a plurality of full connection layers of the decoder to obtain the decoded value;
wherein the solutionThe code formula is:wherein (1)>Is said corrected fly-height-fly-speed interaction feature vector,/>Is the decoded value,/->Is a weight matrix, < >>Is a bias vector, ++>Representing matrix multiplication->To activate the function.
5. The power equipment self-adaptive control method for the supersonic unmanned aerial vehicle is characterized by comprising the following steps of:
acquiring flying height values and flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points in a preset time period; and
adaptively adjusting a thrust output value of the supersonic unmanned aerial vehicle based on the flying height values and flying speed values of the supersonic unmanned aerial vehicle at a plurality of preset time points;
wherein, based on the flying height value and the flying speed value of the supersonic unmanned aerial vehicle at a plurality of preset time points, the thrust output value of the supersonic unmanned aerial vehicle is adaptively adjusted, comprising:
respectively arranging the flying height values and flying speed values of the controlled supersonic unmanned aerial vehicle at a plurality of preset time points into flying height time sequence input vectors and flying speed time sequence input vectors according to time dimensions;
extracting a flight altitude time sequence feature vector and a flight speed time sequence feature vector from the flight altitude time sequence input vector and the flight speed time sequence input vector;
performing feature interaction on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector;
generating a recommended thrust output value of the aircraft engine based on the altitude-speed interaction feature vector; and
based on the recommended thrust output value of the aircraft engine, adaptively adjusting the thrust output value of the supersonic unmanned aerial vehicle;
wherein extracting a flight level timing feature vector and a flight speed timing feature vector from the flight level timing input vector and the flight speed timing input vector comprises:
the flying height time sequence input vector passes through a flying height time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying height time sequence feature vector; and
the flying speed time sequence input vector passes through a flying speed time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain the flying speed time sequence feature vector;
the feature interaction is performed on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a flying height-flying speed interaction feature vector, and the feature interaction method comprises the following steps:
performing feature interaction between the flying height time sequence feature vector and the flying speed time sequence feature vector by using a cascading function to obtain a flying height-flying speed interaction feature vector;
wherein the cascade function is formulated as:wherein,and->All representing a point convolution of the input,/->To activate the function +.>The operation of the splice is indicated and,characteristic values representing respective positions in the flying height timing characteristic vector, +.>Characteristic values representing the positions in the time sequence characteristic vector of the flying speed;
wherein generating a recommended thrust output value of the aircraft engine based on the altitude-speed interaction feature vector comprises:
carrying out forward propagation information retention fusion on the flying height time sequence feature vector and the flying speed time sequence feature vector to obtain a correction feature vector;
based on the correction feature vector, carrying out feature distribution correction on the flying height-flying speed interaction feature vector to obtain a correction flying height-flying speed interaction feature vector; and
and carrying out decoding regression on the corrected flying height-flying speed interaction characteristic vector through a decoder to obtain a decoding value, wherein the decoding value is used for representing the recommended thrust output value of the aircraft engine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310917386.1A CN116880198B (en) | 2023-07-25 | 2023-07-25 | Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310917386.1A CN116880198B (en) | 2023-07-25 | 2023-07-25 | Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116880198A CN116880198A (en) | 2023-10-13 |
| CN116880198B true CN116880198B (en) | 2024-03-12 |
Family
ID=88271302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310917386.1A Active CN116880198B (en) | 2023-07-25 | 2023-07-25 | Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116880198B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118466368B (en) * | 2024-07-09 | 2024-10-18 | 克拉玛依市远山石油科技有限公司 | Working state remote monitoring system and method for unmanned aerial vehicle aeromagnetic equipment |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101264795A (en) * | 2007-03-16 | 2008-09-17 | 王存孝 | Superaltitude low speed positioning flight pilotless aircraft |
| RU2579796C1 (en) * | 2015-02-06 | 2016-04-10 | Открытое акционерное общество Московский научно-производственный комплекс "Авионика" имени О.В. Успенского (ОАО МНПК "Авионика") | Method of determining aircraft engine thrust |
| RU2601367C1 (en) * | 2015-06-18 | 2016-11-10 | Открытое акционерное общество Московский научно-производственный комплекс "Авионика" имени О.В. Успенского (ОАО МНПК "Авионика") | Method of aircraft engine thrust determining at operating mode change and device for its implementation |
| CN109358634A (en) * | 2018-11-20 | 2019-02-19 | 南京航空航天大学 | A kind of hypersonic aircraft Robust Adaptive Control method |
| CN109426146A (en) * | 2017-08-23 | 2019-03-05 | 中国空气动力研究与发展中心计算空气动力研究所 | The high-order Non-Singular Terminal Sliding Mode control method of hypersonic aircraft |
| CN110231774A (en) * | 2019-04-08 | 2019-09-13 | 天津大学 | Disturbance-observer becomes air intake duct hypersonic aircraft fuzzy coordinated control method |
| CN110488609A (en) * | 2019-09-02 | 2019-11-22 | 南京航空航天大学 | A kind of aero-engine direct Thrust inverse control method and device |
| CN114658563A (en) * | 2022-04-15 | 2022-06-24 | 中国航空发动机研究院 | Turbine-based combined cycle engine, aircraft and control method |
| CN114741965A (en) * | 2022-04-18 | 2022-07-12 | 中国航发沈阳发动机研究所 | Hybrid power aeroengine and optimization design method thereof |
| CN116088566A (en) * | 2022-12-19 | 2023-05-09 | 中国航天空气动力技术研究院 | Unmanned aerial vehicle ground speed self-adaptive control method, controller and unmanned aerial vehicle |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3022340B1 (en) * | 2014-06-16 | 2016-07-15 | Thales Sa | METHOD AND DEVICE FOR DETERMINING AN AIRCRAFT CONTROL INSTRUCTION, COMPUTER PROGRAM PRODUCT AND ASSOCIATED AIRCRAFT |
| US9944388B2 (en) * | 2015-08-27 | 2018-04-17 | Sikorsky Aircraft Corporation | Rotorcraft state control |
| FR3089501B1 (en) * | 2018-12-07 | 2021-09-17 | Safran Aircraft Engines | COMPUTER ENVIRONMENT SYSTEM FOR AIRCRAFT ENGINE MONITORING |
-
2023
- 2023-07-25 CN CN202310917386.1A patent/CN116880198B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101264795A (en) * | 2007-03-16 | 2008-09-17 | 王存孝 | Superaltitude low speed positioning flight pilotless aircraft |
| RU2579796C1 (en) * | 2015-02-06 | 2016-04-10 | Открытое акционерное общество Московский научно-производственный комплекс "Авионика" имени О.В. Успенского (ОАО МНПК "Авионика") | Method of determining aircraft engine thrust |
| RU2601367C1 (en) * | 2015-06-18 | 2016-11-10 | Открытое акционерное общество Московский научно-производственный комплекс "Авионика" имени О.В. Успенского (ОАО МНПК "Авионика") | Method of aircraft engine thrust determining at operating mode change and device for its implementation |
| CN109426146A (en) * | 2017-08-23 | 2019-03-05 | 中国空气动力研究与发展中心计算空气动力研究所 | The high-order Non-Singular Terminal Sliding Mode control method of hypersonic aircraft |
| CN109358634A (en) * | 2018-11-20 | 2019-02-19 | 南京航空航天大学 | A kind of hypersonic aircraft Robust Adaptive Control method |
| CN110231774A (en) * | 2019-04-08 | 2019-09-13 | 天津大学 | Disturbance-observer becomes air intake duct hypersonic aircraft fuzzy coordinated control method |
| CN110488609A (en) * | 2019-09-02 | 2019-11-22 | 南京航空航天大学 | A kind of aero-engine direct Thrust inverse control method and device |
| CN114658563A (en) * | 2022-04-15 | 2022-06-24 | 中国航空发动机研究院 | Turbine-based combined cycle engine, aircraft and control method |
| CN114741965A (en) * | 2022-04-18 | 2022-07-12 | 中国航发沈阳发动机研究所 | Hybrid power aeroengine and optimization design method thereof |
| CN116088566A (en) * | 2022-12-19 | 2023-05-09 | 中国航天空气动力技术研究院 | Unmanned aerial vehicle ground speed self-adaptive control method, controller and unmanned aerial vehicle |
Non-Patent Citations (2)
| Title |
|---|
| Double actuated planar single legged hopping robot forward speed and hopping height control;Ahmed A.A. Fahmy et al.;2013 5th International Conference on Modelling, Identification and Control (ICMIC);20131024;全文 * |
| 中空长航时无人机恒速螺旋桨与发动机匹配研究;温占永 等;航空工程进展;20201031;第11卷(第5期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116880198A (en) | 2023-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wicker et al. | Robustness of 3d deep learning in an adversarial setting | |
| CN108009525B (en) | A kind of specific objective recognition methods over the ground of the unmanned plane based on convolutional neural networks | |
| JP6924517B2 (en) | How to recognize faces using multiple patch combinations of deep neural network infrastructure to improve fault tolerance and fracture robustness in extreme situations | |
| Mackowiak et al. | Generative classifiers as a basis for trustworthy image classification | |
| CN116880198B (en) | Power equipment self-adaptive control system and method for supersonic unmanned aerial vehicle | |
| Akan et al. | Stretchbev: Stretching future instance prediction spatially and temporally | |
| CN109559315B (en) | Water surface segmentation method based on multipath deep neural network | |
| CN107423747A (en) | A kind of conspicuousness object detection method based on depth convolutional network | |
| CN111368935A (en) | SAR time-sensitive target sample augmentation method based on generation countermeasure network | |
| CN116071701A (en) | YOLOv5 pedestrian detection method based on attention mechanism and GSConv | |
| CN114529765A (en) | Data processing method, data processing equipment and computer readable storage medium | |
| CN115063672A (en) | Small target detection method based on improved YOLOX model | |
| Wang et al. | Rank-based filter pruning for real-time uav tracking | |
| CN112673384A (en) | Apparatus and method for training amplification discriminators | |
| CN111104855B (en) | Workflow identification method based on time sequence behavior detection | |
| KR102210404B1 (en) | Location information extraction device and method | |
| CN109145738B (en) | Dynamic video segmentation method based on weighted non-convex regularization and iterative re-constrained low-rank representation | |
| CN117150272A (en) | Unmanned aerial vehicle real-time track prediction method and system based on artificial intelligence | |
| CN112487927B (en) | Method and system for realizing indoor scene recognition based on object associated attention | |
| CN115994242A (en) | Image retrieval method, device, equipment and storage medium | |
| CN115631444A (en) | Unmanned aerial vehicle aerial image target detection algorithm | |
| CN114241278A (en) | Multi-branch pedestrian re-identification method and system | |
| CN111401155B (en) | Image recognition method of residual error neural network based on implicit Euler jump connection | |
| CN110188692B (en) | Enhanced cyclic cascading method for effective target rapid identification | |
| CN116449860A (en) | Unmanned aerial vehicle rapid path planning method and system based on deep 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |