CN116484177B - Motion-induced noise prediction elimination method for electromagnetic detection of flight platform - Google Patents
Motion-induced noise prediction elimination method for electromagnetic detection of flight platform Download PDFInfo
- Publication number
- CN116484177B CN116484177B CN202310485740.8A CN202310485740A CN116484177B CN 116484177 B CN116484177 B CN 116484177B CN 202310485740 A CN202310485740 A CN 202310485740A CN 116484177 B CN116484177 B CN 116484177B
- Authority
- CN
- China
- Prior art keywords
- motion
- data
- noise
- flight
- motion state
- 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 54
- 238000001514 detection method Methods 0.000 title claims abstract description 23
- 230000008030 elimination Effects 0.000 title claims abstract description 11
- 238000003379 elimination reaction Methods 0.000 title claims abstract description 11
- 238000013016 damping Methods 0.000 claims abstract description 30
- 238000003062 neural network model Methods 0.000 claims abstract description 18
- 238000012360 testing method Methods 0.000 claims abstract description 14
- 238000013507 mapping Methods 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims abstract description 10
- 230000001133 acceleration Effects 0.000 claims description 19
- 238000005070 sampling Methods 0.000 claims description 6
- 230000005358 geomagnetic field Effects 0.000 claims description 5
- 230000035939 shock Effects 0.000 claims description 5
- 230000024703 flight behavior Effects 0.000 claims 2
- 238000005259 measurement Methods 0.000 abstract description 7
- 238000013528 artificial neural network Methods 0.000 description 18
- 230000006698 induction Effects 0.000 description 15
- 230000006870 function Effects 0.000 description 13
- 230000008569 process Effects 0.000 description 9
- 230000008859 change Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000013459 approach Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 5
- 210000002569 neuron Anatomy 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Navigation (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention provides a motion-induced noise prediction elimination method for electromagnetic detection of a flying platform, which comprises the following steps: performing a test flight of pure noise acquisition for one time before the formal survey line flight, and acquiring a group of motion state data and corresponding pure noise data; training a wavelet neural network model based on the experimental flight motion state data and the pure noise data; acquiring motion state data and receiving data of the flight of the survey line; mapping the motion state data based on the small neural network model, and predicting motion-induced noise; and removing the predicted motion-induced noise from the received data to complete the electromagnetic method motion-induced noise elimination based on the flying platform. The method can be used for predicting the motion-induced noise by directly using the motion state measurement data, can be used for a half aviation electromagnetic method and a full aviation electromagnetic method, can remove the same-frequency motion-induced noise with the same frequency as the effective signal, and can predict and eliminate the motion-induced noise of a flight platform with a damping device.
Description
Technical Field
The invention belongs to the technical field of electromagnetic detection based on a flight platform, and particularly relates to a motion state data prediction and motion induction noise elimination technology which can be used for a full-aviation electromagnetic method and a half-aviation electromagnetic method.
Background
The artificial source electromagnetic detection technology is an important branch of geophysical electromagnetic detection, and can be divided into ground, semi-aviation and full-aviation electromagnetic detection methods according to different construction spaces, wherein the semi-aviation and aviation electromagnetic methods belong to electromagnetic detection methods based on flight platforms, and an induction magnetic field in the air is measured through flight platform carrying receiving systems such as helicopters, airships and unmanned aerial vehicles. The semi-aviation electromagnetic detection method excites an underground abnormal body through a ground paved emission source. The full aviation electromagnetic method is characterized in that a transmitting source and a receiving device are carried on a flying platform together, and signals are transmitted and received in the air. Compared with a ground electromagnetic method, the electromagnetic detection system based on the flight platform is flexible in maneuvering and strong in terrain adaptation capability, can overcome the geological resource exploration difficulty under the condition of complex terrains in China, and has the advantage of high-depth rapid detection.
Motion-induced noise is the most influential noise type in all aircraft platform-based electromagnetic detection methods. In the flight process of the receiving coil, the receiving coil is influenced by wind speed, wind power and the flight state of the aircraft, the change of the attitude angle can be generated, the change of the effective area of the receiving coil is caused, and therefore the receiving coil cuts the geomagnetic field to cause the change of the induced magnetic flux. The geomagnetic field has the intensity of 6×10 -5 tesla, which is 10 ten thousand times stronger than the magnetic field intensity of the electromagnetic detection system based on the flying platform, so that the change of any angle of the receiving coil can cause stronger change of induced electromotive force, and motion-induced noise is introduced into the received signal. The motion-induced noise reduces the signal-to-noise ratio and greatly influences the detection depth and detection accuracy of the system.
When a helicopter or an unmanned plane is used as a carrying platform, the rotation frequency of a propeller is 350Hz-400Hz, vibration of the frequency can be transmitted to a magnetic field sensor through a rope system, vibration interference is introduced, meanwhile, no matter the helicopter, the unmanned plane or an airship is used as the carrying platform, abrupt changes of the terrain, the wind speed and the wind direction or abrupt airflow impact are caused, the change of the motion state of the flying platform is also transmitted to the magnetic field sensor through the rope system, and at the moment, the receiving coil can also cause high-frequency vibration due to the impact airflow disturbance. Therefore, in general, the system is designed to add a damping structure to the receiving coil to reduce the vibration frequency of the coil, which makes the motion state of the coil itself undetectable.
Research into motion-induced noise removal techniques can be broadly divided into three approaches.
The first approach is to design correction factors, filter coefficients or directly use measured attitude data to calculate and remove motion-induced noise by observing the flight process according to the generation mechanism of the motion-induced noise. The space-sky yard Liu Fubo of the department of Chinese, etc., predicts the motion-induced noise by directly using the attitude measurement data through a rotation matrix method. The method has the advantages of low observation precision and limited inhibition effect on motion-induced noise when directly observing the flight process. The calculation and removal of motion induced noise using attitude data is a good idea, however, this approach would not be usable in systems with shock absorbing devices.
The second approach is to use signal-to-noise separation and data reconstruction for motion-induced noise removal based on the characteristics of the motion-induced noise. Representative works are: buselli et al propose to use a high pass filter to remove motion-induced noise in view of the low frequency characteristics of avionic motion-induced noise; li Nan performing noise removal by using a wavelet thresholding method by utilizing wavelet decomposition; zhu Kaiguang and the like propose to use principal component analysis to denoise according to the correlation relation between noise and signals; liu Fubo, etc., propose an empirical mode decomposition method in consideration of the periodic characteristics of the effective signals; li Yuan and the like are combined with the characteristics of periodicity of the effective signal and low noise frequency, high-order empirical mode quantity obtained by empirical mode decomposition is reconstructed and subjected to high-pass filtering, and then the signal is reconstructed together with the low-order empirical mode so as to save the effective signal as much as possible.
However, in the case where the signal-to-noise distinction is not strong, there is a tendency that the signal-to-noise cannot be completely separated, which results in the following: 1) Part of the effective signal is damaged with the removal of motion-induced noise; 2) The effective signal is kept complete, and the motion-induced noise is not removed cleanly; 3) The motion-induced noise is not removed cleanly and the effective signal is damaged. Typically, such methods tend to be ineffective when motion-induced noise is co-frequency with the effective signal.
The third approach is to fit and extend the late signal to a full-wave half-cycle signal by treating it as pure motion-induced noise data, and then subtracting the fitted noise data from the original signal. According to the method, after the emission signal is considered to be off, the secondary field signal is gradually weakened to be close to zero, and the secondary field signal is annihilated as motion induction noise, so that late data after annihilation points can be used as pure motion induction noise, and the motion induction noise of the whole half period can be calculated through fitting expansion. However, motion-induced noise has the characteristic of non-stability, late-stage signals are not easy to expand to the whole half period, the late-stage data volume after annihilation points is small, the noise source is complex and extremely unstable, the fitting process is inevitably unstable, and the fitting is severely distorted.
The motion-induced noise is caused by the motion state change of the receiving coil, so that the motion-induced noise can be calculated and eliminated by measuring the motion state of the receiving coil, however, the sampling frequency limitation of the motion state measuring sensor makes the high-frequency motion-induced noise impossible to predict, and the noise is very strong and mixed with the effective signal frequency, so that in the electromagnetic system based on the flying platform, a damping structure is generally added on the receiving coil to reduce the vibration frequency of the coil, which makes the motion state of the coil not measurable. At this time, the motion state sensor is installed outside the shock absorbing device, and the measured motion state is not the motion state of the coil, however, it is still closely related to the motion state of the receiving coil. Meanwhile, the motion state of the receiving coil is related to wind speed, the whole system structure, the flying condition and the like, the motion process of the receiving coil is limited by the whole system structure, and the receiving coil and the whole system have certain internal relation and constraint conditions, and the internal relation and constraint conditions are the same for the same avionics system. However, these inherent relationships and constraints are too complex to be expressed by formula derivation. How to fully use the measured motion state, solve the motion-induced noise through the inherent relation and constraint condition, and complete the prediction and elimination of the motion-induced noise of the electromagnetic detection system with the damping device based on the flying platform is the main research content of the invention.
The electromagnetic method based on the flight platform comprises a time domain full-aviation electromagnetic method, a frequency domain full-aviation electromagnetic method and a half-aviation electromagnetic method. Fig. 1 is a typical device in a time domain all-air electromagnetic method: a concentric time domain airborne electromagnetic device. The nacelle of the concentric time domain aviation electromagnetic device mainly comprises a transmitting coil, a compensating coil and a receiving coil, wherein the three coils are connected by ropes, and the motion states of the three coils are not identical. The sensors for measuring motion state data (three-axis attitude, three-axis acceleration, three-component speed sensors) of the concentric time domain electromagnetic system are installed outside the transmitting coil with the damping device, and also installed on the compensating coil (shown in fig. 4) or the transmitting coil, and the method of the patent can be used under the above conditions. Fig. 2 is a typical device in the frequency domain full-air electromagnetic method, where sensors for measuring motion state data are often mounted in the nacelle. Fig. 3 is a schematic diagram of a typical device for a semi-aeronautical transient electromagnetic method, which receives signals only in the air, and a sensor for measuring movement state data is mounted on a receiving device with a shock absorbing structure.
Disclosure of Invention
In order to solve the technical problems, the invention provides a technology for a motion-induced noise prediction elimination method for electromagnetic detection of a flight platform, which predicts and eliminates motion-induced noise by using motion state data through a wavelet neural network, wherein the wavelet neural network method is used for establishing a set of complex nonlinear mapping networks to describe the inherent relation and constraint conditions of motion of a receiving coil and a system integral structure, flight condition, flight acceleration and the like, and the network can be used for simulating the damping process of a damping device in a system with the damping device outside a transmitting coil.
A motion-induced noise prediction cancellation method for electromagnetic detection of a flying platform, comprising:
Before the formal survey line flies, pure noise acquisition test flying of closing emission for one time is carried out according to a normal survey line flying mode, the flying duration is not less than half an hour, and a group of motion state data and pure noise data are acquired;
Training a wavelet neural network model based on the experimental flight motion state data and the pure noise data;
Acquiring motion state data and receiving data of the flight of the survey line;
Mapping the motion state data based on the wavelet neural network model, and predicting motion-induced noise;
acquiring motion state data and a receiving signal of the flight of the survey line;
Mapping the motion state data based on the wavelet neural network model, and predicting motion-induced noise;
and removing the predicted motion-induced noise from the received data to complete the electromagnetic method motion-induced noise elimination based on the flying platform.
Optionally, the flying platform electromagnetic method is a half aviation electromagnetic method or a full aviation electromagnetic method; the flying platform lifting tool is a helicopter, an airship or an unmanned plane; the motion state data comprise a three-axis gesture, a three-axis acceleration and a three-component speed; the motion-induced noise is noise caused by the cutting of the geomagnetic field by the receiving coil.
Optionally, in the time domain aviation electromagnetic system, a sensor for detecting motion state data is arranged on the compensation coil, the receiving coil or a damping device for wrapping the receiving coil, and the receiving coil and the compensation coil are in hard connection or soft connection, wherein the hard connection is a hard bracket, and the soft connection is a soft rope; in the frequency domain aviation electromagnetic system, a sensor for detecting motion state data is arranged on the integral nacelle; in semi-aviation electromagnetic systems, sensors that detect motion state data are mounted on a receiving coil or a shock absorbing device that wraps the receiving coil.
Alternatively, the test flight is a straight flight of not less than half an hour in accordance with the normal line flight mode, which differs from the normal line flight in that the transmitting device is turned off in the test flight, and thus no effective signal is input in the receiving device.
Optionally, the test flight motion state data and the pure noise data are used, and when the wavelet neural network model is trained, the output is pure noise data y= [ Y 1,Y2,…Yi,…,Yq ], wherein Y i is the pure noise data at time t i.
When training a wavelet neural network model by using test flight motion state data and pure noise data, the input is motion state data:
X＝[X1,X2,…Xi,…,Xq]
Where θ is the attitude angle, a is the acceleration, v is the velocity, and their first subscript indicates the axial direction, e.g., for attitude angle θ, the first subscript indicates 1 the attitude angle of rotation about the x-axis, 2 the attitude angle of rotation about the y-axis, and 3 the attitude angle of rotation about the z-axis. m, n, p represent the sampling points of the attitude angle, the acceleration and the speed from k seconds before the t moment to k seconds after the t moment. X i represents a motion state parameter related to noise Y i at time t i.
Compared with the prior art, the invention has the following advantages and technical effects:
For the same set of avionics system, the damping properties of the damping device are the same, the internal relations and constraint conditions of the motion of the receiving coil in the damping device and the system structure, the flight condition, the flight acceleration and the like are the same, a set of complex nonlinear mapping networks are established by using the wavelet neural network to describe the internal relations and constraint conditions of the motion of the receiving coil and the system integral structure, the flight condition, the flight acceleration and the like, and the damping process of the damping device can be simulated in the system with the damping device, so that a mapping relation (wavelet neural network) reflecting the motion state of the system to the motion induced noise is finally obtained. The invention directly uses the motion state measurement data to map the motion induction noise, and can remove the same-frequency motion induction noise with the same frequency as the effective signal. The prediction and elimination of motion-induced noise can be performed on a system having a damping device.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:
Fig. 1 is a schematic diagram of a detection principle of a central loop type device commonly used in a time domain all-aviation electromagnetic detection device in the background art;
FIG. 2 is a schematic diagram of a typical apparatus in a frequency domain full aviation electromagnetic detection method in the background art;
FIG. 3 is a schematic diagram of a typical apparatus in a semi-airborne electromagnetic detection method in the background;
FIG. 4 is a schematic view of a transmitting-receiving plane of a conventional receiving coil with a damping device mounted thereon for a center loop device according to the prior art;
FIG. 5 is a schematic diagram of the basic structure of a neural network according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a single neuron model according to an embodiment of the present invention;
fig. 7 is a schematic flow chart of an airborne electromagnetic motion-induced noise cancellation method based on a wavelet neural network according to an embodiment of the invention.
Detailed Description
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
The artificial neural network is used as one of machine learning algorithms and is a research hotspot in the field of artificial intelligence in the 80 th century. The algorithm establishes networks with different connection structures in a mode of simulating brain to information processing, and the networks become complex nonlinear mapping functions capable of realizing a certain specific function after specific training. The basic structure of the neural network is shown in fig. 5, and the neural network consists of an input layer, a hidden layer and an output layer, and reflects a mapping relation between input data and output data. The hidden layer is composed of a series of neurons (as shown in fig. 6, the activation function g (r) is replaced by a wavelet basis function ψ), the output of the neurons is obtained by respectively carrying out weighted summation on all the inputs of the neurons according to corresponding weights, then carrying out addition with bias to obtain a state function r, and finally carrying out nonlinear activation on the state function r through a certain specific activation function g (r).
The Wavelet Neural Network (WNN) is a neural network structure obtained by replacing the activation function of each neuron of the neural network by a wavelet function, and respectively replacing the weight from the corresponding input layer to the hidden layer and the threshold value of the hidden layer by a scale expansion factor and a time shift factor of the wavelet function. The WNN integrates the advantages of the artificial neural network and the wavelet analysis, so that the network convergence speed is high, the local optimization is avoided, and the time-frequency local analysis is realized.
For the damping process with the damping structure, the intrinsic relation and constraint conditions of the motion of the receiving coil in the damping structure, the system structure, the flight condition, the flight acceleration and the like are used for obtaining a set of neural network parameters by measuring the motion state data of the system such as the three-axis posture, the three-axis acceleration, the three-component speed and the like as network input and training the model without developing the pure motion induction noise received by the corresponding period as output. And finally, mapping and predicting motion-induced noise through the neural network by using the motion state data acquired in the acquisition stage of the measuring line, and subtracting the motion-induced noise from the measured measuring line data to obtain a denoising signal. According to the denoising concept, this embodiment proposes a four-stage workflow (as shown in fig. 7), and uses a wavelet neural network to predict and eliminate motion-induced noise.
1. And (5) data acquisition. Before the formal survey line flight, performing a pure noise acquisition test flight of closing emission for one time according to a normal survey line flight mode, wherein the flight is a straight line flight with the duration not less than half an hour, so that a group of motion state data and receiving data are acquired; the received data is pure noise data, and the received data does not contain effective signals, only contains motion induction noise caused by the geomagnetic field of coil cutting.
After pure noise data are collected, the emission is started, normal line flight is carried out according to the designed flight line, and motion state data and receiving data are collected.
2. And constructing a wavelet neural network model, taking the motion state data acquired in the test flight as input, and taking the pure noise data as output to execute WNN training.
3. And using the trained WNN, taking the motion state data acquired by the line flight as input, and predicting the motion-induced noise corresponding to the earth response and the motion-induced noise in the line data.
4. The predicted motion-induced noise is subtracted from the corresponding line data.
In the data acquisition stage, in order to better utilize the inherent relation between the motion state of the transmitting plane and the motion-induced noise, so as to calculate the motion-induced noise data and separate effective signals, we install sensors on the transmitting coil plane as much as possible to obtain more relevant motion state data: three-axis attitude sensor, three-axis acceleration sensor, three-axis speed sensor.
The sample set training of stage 2 is considered to be a critical step:
(1) Constructing the wavelet neural network model: setting the number of elements contained in each layer of the wavelet neural network model according to the characteristics of signals;
(2) Initializing the wavelet neural network model: randomly initializing scale factors and translation factors, determining input and output connection weights of a wavelet function, and setting a learning rate;
(3) Prediction output: acquiring an input value of each node of a hidden layer in the wavelet neural network model; inserting the input value of each node of the hidden layer into a parent wavelet basis function, and calculating the output value of the hidden layer; calculating an output value of the wavelet neural network model by using the output connection weight;
(4) Updating parameters: calculating network prediction deviation; correcting scale factors and translation factors of wavelet functions and weights of neural network connection based on the network prediction bias; and (5) evaluating the quality of network output and finishing iteration.
For our input data, since the motion-induced noise is related to the change of the posture, the damping device has damping motion, and therefore, the present embodiment must consider the motion state data between k seconds before t and k seconds after t when predicting the motion-induced noise at t.
The output of this embodiment is pure noise data y= [ Y 1,Y2,…Yi,…,Yq ], where Y i is pure noise data at time t i;
The input of this embodiment is motion state data:
X＝[X1,X2,…Xi,…,xq]
Where θ is the attitude angle, a is the acceleration, v is the velocity, and their first subscript indicates the axial direction, e.g., for attitude angle θ, the first subscript indicates 1 the attitude angle of rotation about the x-axis, 2 the attitude angle of rotation about the y-axis, and 3 the attitude angle of rotation about the z-axis. m, n, p represent the sampling points of the attitude angle, the acceleration and the speed between k seconds before the t moment and k seconds after the t moment, if the sampling rates of the attitude angle, the acceleration and the speed are the same, m, n and p are equal, and if the sampling rates are different, the three are different. X i represents a motion state parameter related to noise Y i at time t i.
The embodiment proposes the operation of closing the emission signal before measuring the survey line, collecting a group of motion state data (three-axis gesture, three-axis acceleration, three-component speed) of the system and corresponding pure noise data under normal flight operation, wherein the effective signal and the motion-induced noise are not coupled and have additive relation, so that the mapping relation between the motion state measurement data and the motion-induced noise under the operation can be used in the survey line data containing the effective signal.
The mapping relation between the plane motion state measurement data of the transmitting coil and the corresponding motion induction noise data comprises a damping process of a damping device, and an intrinsic relation and constraint conditions of motion and system structure, flight condition, flight acceleration and the like of a receiving coil in the damping device.
The embodiment proposes that the motion state measurement data of the measuring line is used as input, the trained wavelet neural network model is used for predicting the motion induction noise, and the motion induction noise is subtracted from the data of the measuring line, and the method is effective for the same-frequency motion induction noise with the same frequency as the effective signal.
The advantages of this embodiment are:
For the same set of avionic systems, the damping properties of the damping devices are the same, the motion of the receiving coils in the damping devices is the same as the intrinsic relation and constraint conditions of the system structure, the flight condition, the flight acceleration and the like, the intrinsic relation and constraint conditions are trained and mapped by using the wavelet neural network through the method of the patent, the intrinsic relation and constraint conditions can be fully utilized for denoising, and the denoising precision is improved. The method directly uses the motion state measurement data to map the motion induction noise, and can remove the same-frequency motion induction noise with the same frequency as the effective signal. The prediction and elimination of motion-induced noise can be performed on a system having a damping device.
The present application is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present application are intended to be included in the scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims (3)
1. A motion-induced noise prediction cancellation method for electromagnetic detection of a flying platform, comprising:
Performing a pure noise acquisition test flight before the formal survey line flight, and acquiring a group of motion state data and pure noise data;
The test flight behavior is a straight line flight which is carried out according to a normal line flight mode and is not less than half an hour, and the test flight behavior is different from the normal line flight in that a transmitting device is closed in the test flight, and no effective signal is input in a receiving device;
Training a wavelet neural network model based on the experimental flight motion state data and the pure noise data;
When the test flight motion state data and the pure noise data are used for training the wavelet neural network model, the output is the pure noise data Y= [ Y 1,Y2,…Yi,…,Yq ], wherein Y i is the pure noise data at the moment t i;
when training a wavelet neural network model by using test flight motion state data and pure noise data, the input is motion state data:
X＝[X1,X2,…Xi,…,Xq]
Where θ is the attitude angle, a is the acceleration, v is the velocity, their first subscript indicates the axial direction, for example, for attitude angle θ, first subscript 1 indicates the attitude angle of rotation about the X-axis, 2 indicates the attitude angle of rotation about the Y-axis, 3 indicates the attitude angle of rotation about the z-axis, m, n, p indicates the number of sampling points of the attitude angle, acceleration, velocity between k seconds before t and k seconds after t, and X i indicates the motion state parameter related to noise Y i at time t i;
Acquiring motion state data and receiving data of the flight of the survey line;
Mapping the motion state data based on the wavelet neural network model, and predicting motion-induced noise;
and removing the predicted motion-induced noise from the received data to complete the electromagnetic method motion-induced noise elimination based on the flying platform.
2. The motion-induced noise prediction cancellation method for electromagnetic surveying of a flying platform of claim 1, wherein: the flight platform electromagnetic method is a half aviation electromagnetic method or a full aviation electromagnetic method; the flying platform lifting tool is a helicopter, an airship or an unmanned plane; the motion state data comprise a three-axis gesture, a three-axis acceleration and a three-component speed; the motion-induced noise is noise caused by the cutting of the geomagnetic field by the receiving coil.
3. The motion-induced noise prediction cancellation method for electromagnetic surveying of a flying platform of claim 1, wherein: in the time domain aviation electromagnetic system, a sensor for detecting motion state data is arranged on a compensation coil, a receiving coil or a damping device for wrapping the receiving coil, and the receiving coil and the compensation coil are in hard connection or soft connection, wherein the hard connection is a hard bracket, and the soft connection is a soft rope; in the frequency domain aviation electromagnetic system, a sensor for detecting motion state data is arranged on the integral nacelle; in semi-aviation electromagnetic systems, sensors that detect motion state data are mounted on a receiving coil or a shock absorbing device that wraps the receiving coil.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310485740.8A CN116484177B (en) | 2023-05-04 | 2023-05-04 | Motion-induced noise prediction elimination method for electromagnetic detection of flight platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310485740.8A CN116484177B (en) | 2023-05-04 | 2023-05-04 | Motion-induced noise prediction elimination method for electromagnetic detection of flight platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116484177A CN116484177A (en) | 2023-07-25 |
| CN116484177B true CN116484177B (en) | 2024-05-28 |
Family
ID=87213628
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310485740.8A Active CN116484177B (en) | 2023-05-04 | 2023-05-04 | Motion-induced noise prediction elimination method for electromagnetic detection of flight platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116484177B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116628429B (en) * | 2023-07-26 | 2023-10-10 | 青岛远度智能科技有限公司 | Intelligent control method for stable lifting of unmanned aerial vehicle |
| CN117110953B (en) | 2023-10-24 | 2024-01-02 | 中国科学院地质与地球物理研究所 | Mobile mineral resource rapid investigation method, device and application thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105534517A (en) * | 2016-02-29 | 2016-05-04 | 浙江铭众科技有限公司 | Method for removing vehicle motion noise in three-lead electrocardiosignal |
| CN112068212A (en) * | 2020-07-30 | 2020-12-11 | 成都理工大学 | Analysis and interpretation method for semi-aviation time domain electromagnetic detection data of unmanned aerial vehicle |
| CN113128115A (en) * | 2021-04-16 | 2021-07-16 | Oppo广东移动通信有限公司 | Subway running state prediction and model training method and device and storage medium |
| CN113341469A (en) * | 2021-06-02 | 2021-09-03 | 山东大学 | Semi-aviation transient electromagnetic data correction method and system |
| WO2022126396A1 (en) * | 2020-12-15 | 2022-06-23 | 深圳市大疆创新科技有限公司 | State information determination method and apparatus, and platform and computer-readable storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019232539A1 (en) * | 2018-06-01 | 2019-12-05 | New York University | System, method and computer-accessible medium for facilitating noise removal in magnetic resonance imaging |
| US10877178B2 (en) * | 2019-05-03 | 2020-12-29 | Institute Of Geology And Geophysics, Chinese Academy Of Sciences | Method for suppressing airborne transient electromagnetic in-band vibration noise |
-
2023
- 2023-05-04 CN CN202310485740.8A patent/CN116484177B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105534517A (en) * | 2016-02-29 | 2016-05-04 | 浙江铭众科技有限公司 | Method for removing vehicle motion noise in three-lead electrocardiosignal |
| CN112068212A (en) * | 2020-07-30 | 2020-12-11 | 成都理工大学 | Analysis and interpretation method for semi-aviation time domain electromagnetic detection data of unmanned aerial vehicle |
| WO2022126396A1 (en) * | 2020-12-15 | 2022-06-23 | 深圳市大疆创新科技有限公司 | State information determination method and apparatus, and platform and computer-readable storage medium |
| CN113128115A (en) * | 2021-04-16 | 2021-07-16 | Oppo广东移动通信有限公司 | Subway running state prediction and model training method and device and storage medium |
| CN113341469A (en) * | 2021-06-02 | 2021-09-03 | 山东大学 | Semi-aviation transient electromagnetic data correction method and system |
Non-Patent Citations (4)
| Title |
|---|
| YiguangChen,etc.Winding and Electromagnetic Analysis for 39- Slot/12-Pole Frameless Permanent Magnet Synchronous Motor.IEEE.2019,全文. * |
| 唐崇凯 ; 曲建岭 ; 高峰 ; 周胜明 ; .基于小波神经网络的飞参数据预处理.数据采集与处理.2012,(第S1期),正文. * |
| 孙珽 ; 徐东星 ; 尹勇 ; 张秀凤 ; 苌占星 ; 叶进 ; .基于VDM与APSO优化极限学习机的船舶运动姿态预报.船舶工程.2019,第41卷(第11期),正文. * |
| 尹雄等.中心回线航空瞬变电磁探测系统误差与噪声的定量分析.《地球物理学进展》.2022,第 第37卷卷(第第2期期),正文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116484177A (en) | 2023-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116484177B (en) | Motion-induced noise prediction elimination method for electromagnetic detection of flight platform | |
| US10877178B2 (en) | Method for suppressing airborne transient electromagnetic in-band vibration noise | |
| CN211741603U (en) | Device for infrared detecting submarine | |
| Wu et al. | The removal of the high-frequency motion-induced noise in helicopter-borne transient electromagnetic data based on wavelet neural network | |
| CN111950503B (en) | Aviation transient electromagnetic data processing method and device and computing equipment | |
| Wu et al. | A deep learning estimation of the earth resistivity model for the airborne transient electromagnetic observation | |
| US8359162B2 (en) | Terrain correction systems | |
| CN108072910B (en) | A kind of distribution magnetic anomaly detection system environment magnetic compensation method | |
| CN109814163B (en) | Method and system for suppressing noise of aeromagnetic tensor data based on extended compensation model | |
| CN112710470B (en) | Self-adaptive wavelet threshold MEMS gyroscope noise reduction method based on maneuver identification | |
| CN110187393B (en) | Aeromagnetic compensation method based on generalized regression neural network | |
| Miao et al. | Magnetic anomaly detection based on fast convergence wavelet artificial neural network in the aeromagnetic field | |
| CN112818876B (en) | Electromagnetic signal extraction and processing method based on deep convolutional neural network | |
| CN111948711B (en) | Method and system for extracting low-frequency part of seismic data by deep learning method | |
| D’Amato et al. | UKF-based fault detection and isolation algorithm for IMU sensors of Unmanned Underwater Vehicles | |
| CN111857176A (en) | GPS unmanned aerial vehicle control method | |
| CN110596774A (en) | Method and device for infrared detection of submarine | |
| CN112327371B (en) | Satellite magnetic field data time-varying background field establishment method based on variational modal decomposition | |
| Hezel | Improving aeromagnetic calibration using artificial neural networks | |
| CN116661010A (en) | Ocean magnetic field detection system and detection method based on unmanned aerial vehicle cluster | |
| CN109885906A (en) | Magnetic resonance sounding signal sparse noise elimination method based on particle swarm optimization | |
| CN114578439A (en) | Magnetic compensation coefficient self-adaptive optimization method and system | |
| Ma et al. | Outlier correction method of telemetry data based on wavelet transformation and Wright criterion | |
| Nelson | Aeromagnetic noise during low-altitude flights over the Scotian Shelf | |
| Al-Haddad et al. | An Intelligent Fault Diagnosis Approach for Multirotor UAVs Based on Deep Neural Network of Multi-Resolution Transform Features, Drones, 2023, 7, 82 |
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 |