CN114323635B - Valve fault state sensing diagnosis method and system based on terahertz radar - Google Patents
Valve fault state sensing diagnosis method and system based on terahertz radar Download PDFInfo
- Publication number
- CN114323635B CN114323635B CN202210018701.2A CN202210018701A CN114323635B CN 114323635 B CN114323635 B CN 114323635B CN 202210018701 A CN202210018701 A CN 202210018701A CN 114323635 B CN114323635 B CN 114323635B
- Authority
- CN
- China
- Prior art keywords
- valve
- signal
- radar
- intermediate frequency
- phase
- 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
Landscapes
- Radar Systems Or Details Thereof (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a valve fault state sensing and diagnosing method and system based on terahertz radar, which utilize the terahertz radar to emit linear frequency modulation continuous waves to a valve which is working, the surrounding environment of the valve and objects; receiving echo signals reflected by the valve and the surrounding environment and objects of the valve by using a terahertz radar; mixing an echo signal with the frequency modulation continuous wave by using the terahertz radar to obtain an intermediate frequency signal; performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal; inputting the phase information of the intermediate frequency signal into a trained deep neural network; and outputting the fault part and the fault degree by using the deep neural network to finish the fault diagnosis of the valve core and the valve rod of the electric reversing valve in the valve. The invention makes up the defect that the traditional method can not distinguish the fault types, and can effectively and accurately identify the fault occurrence degree.
Description
Technical Field
The invention belongs to the technical field of valve fault diagnosis, and particularly relates to a valve fault state sensing and calculating diagnosis method and system based on terahertz radar.
Background
The hydraulic system is widely applied to various rotary machines, and the working environment of a control element in the hydraulic system is complex, so that various faults are easy to occur, and the operation of a power system is influenced. In addition to the oil pollution which accounts for about 70% of the system faults, the main fault components of the hydraulic system are pump components and valve components, and the fault of the valve is not accidental, but is a daily and monthly manifestation of the problem. Taking an electric reversing valve in a valve as an example, the function of the electric reversing valve is realized by the relative movement of a valve core and a valve body, and the communication, the cutting-off and the reversing of hydraulic oil, the pressure unloading and the sequential action control can be realized. The valve core and the valve rod are main moving parts of the reversing valve, the stress condition is complex, and the valve core and the valve rod are vulnerable parts. The hydraulic components such as the valve body and the valve sleeve are rubbed continuously in the using and moving process due to the fact that a hydraulic system is unresponsive and frequently occurs due to the abrasion of the valve core of the electric reversing valve and the pull of the valve rod, so that the shape, the size and the surface quality of the hydraulic components are changed, and the abrasion of the valve core of the electromagnetic reversing valve is increased under the condition that excessive impurities are contained in oil. The valve rod of the electric reversing valve is pulled or deformed, so that the stability of the hydraulic system is reduced, the efficiency is reduced, and the valve core is blocked by mechanical blockage due to the fact that foreign matters and dirt are more likely to enter a gap or a deformed position, so that corresponding fault diagnosis research is needed. The main faults of the electric reversing valve are valve core abrasion and valve rod strain, when the valve core and the valve rod of the electric reversing valve are in fault, vibration parameter indexes such as amplitude, vibration frequency and the like can appear, and the indexes reflect the operation working condition of the machine and can help to carry out fault diagnosis.
For valve failure, the traditional vibration measurement method is as follows: a contact vibration measurement method typified by an acceleration sensor and a strain gauge, and a non-contact vibration measurement method typified by a laser doppler vibrometer and a visual vibration measurement system. However, these conventional methods have certain limitations:
(1) The number of measurement points is huge, a large amount of instruments and equipment are generally required to prepare and arrange lines, time and labor are wasted, and most instruments and equipment are difficult to find suitable measurement points for measurement;
(2) For ultra-low frequency vibration frequencies, for weak amplitudes, measurement is difficult and measurement results are not accurate enough;
(3) For a laser Doppler vibration meter and a visual vibration measuring system, an ideal measuring environment is needed, but most of equipment works in a severe environment and is difficult to implement in an ideal testing environment;
(4) The running state and fault early warning function of all-weather monitoring equipment are difficult to realize, and the testing efficiency is relatively low;
(5) The contact acceleration sensor needs to be adhered to the surface of a measured object, so that the normal operation of a light structure can be influenced, and the deviation between a measurement result and an actual result is larger. In addition, there are also methods of laser measurement vibration, ultrasonic measurement vibration, infrared sensor measurement, microwave radar vibration measurement, etc., which have disadvantages. The laser vibration measurement and the infrared sensor are easy to be interfered by external environments, the ultrasonic measurement vibration is limited by the working distance, the distance resolution capability of the microwave radar is too low, and the extremely fine relative displacement change is difficult to measure due to the high wavelength, so that the device is not suitable for vibration monitoring of micro-moving targets. The traditional common method for diagnosing the valve faults is to detect the vibration signals of the valve by using a sensor, further analyze and process the acquired signals to obtain valve fault sensitive characteristic quantities such as acceleration level, impact energy, impact frequency, acceleration frequency domain signals and the like, compare the valve fault sensitive characteristic quantities with normal conditions or detect the valve fault sensitive characteristic quantities for a long time, and obtain the valve fault state and the fault trend, but cannot accurately detect the fault position and the fault degree is difficult to measure.
In an actual industrial environment, when a machine has a tiny fault, the amplitude change is extremely tiny, and accurate detection and early warning are difficult to realize; modern industrial machines all work all the time, if a power system fails, the whole system can be stopped, so that the real-time detection is very important. At present, the existing automatic vibration detection technology is not widely used because of high cost, difficulty in adapting to all scenes, difficulty in large-scale deployment and the like, the fault detection of valve stems of valve cores is judged by using strain gauge vibration detection, most factories can select manual detection at low cost and moderate detection accuracy, but the all-weather detection, the early prediction and the early warning are difficult to realize, the artificial subjective measurement error is large, the existing fault diagnosis method has strict detection environment requirements, the detection efficiency and the detection accuracy are low, and the fault positions and the fault degree cannot be accurately identified.
Disclosure of Invention
The invention aims to solve the technical problems that the valve fault state sensing and diagnosing method and system based on the terahertz radar solve the problems that automatic detection cannot be performed, the detection environment requirement is strict, the detection efficiency and accuracy are low, and the fault occurrence part and the fault occurrence degree cannot be identified in the prior art.
The invention adopts the following technical scheme:
a valve fault state sensing diagnosis method based on terahertz radar comprises the following steps:
s1, transmitting linear frequency modulation continuous waves to a valve which is working by using a terahertz radar and an object around the valve;
s2, receiving echo signals reflected by the valve and the surrounding environment and objects of the valve by using a terahertz radar;
s3, mixing the echo signal received in the step S2 with the frequency modulation continuous wave transmitted in the step S1 by using the terahertz radar to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal;
s4, inputting the phase signal of the intermediate frequency signal obtained in the step S3 into a trained deep neural network;
and S5, outputting a fault part and the fault degree by using the deep neural network in the step S4, and completing the fault diagnosis of the valve core and valve rod of the electric reversing valve in the valve.
Specifically, in step S1, the transmitted chirped continuous wave S (t) is specifically:
wherein a is 0 For the amplitude of the transmitted signal, j is the imaginary part, f c The range of the center frequency of radar emission is terahertz wave band, T is time variable, B is bandwidth of radar, and T is period of linear frequency modulation.
Specifically, in step S2, the echo signal r (t) is specifically:
where k is the attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary part, f c Is the center frequency of radar emission, t is a time variable, t d Is connected withThe time delay of the received signal, B is the bandwidth of the radar and T is the period of the chirp.
Specifically, in step S3, the intermediate frequency signal b (t) within the range of the target distance terahertz radar distance R is:
wherein conj is conjugate operation, s (t) is linear frequency modulation continuous wave emitted by terahertz radar to the valve which is working, the surrounding environment of the valve and objects, r (t) is echo signal, k is attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary part, B is the bandwidth of the radar, R is the distance of the target from the radar, c is the speed of light in vacuum, T is the period of the chirp, T is the time variable, lambda is the wavelength, f b Is the frequency of the intermediate frequency signal b (t), phi b Is the phase of the intermediate frequency signal b (t).
Further, the frequency f of the intermediate frequency signal b (t) b And phase phi b The method comprises the following steps of:
specifically, in step S3, the signal processing on the intermediate frequency signal to obtain the phase signal specifically includes:
s301, performing fast Fourier transform on the intermediate frequency signal, and determining the position of the target according to the peak position obtained by the fast Fourier transform;
s302, carrying out phase solving operation on intermediate frequency signal data belonging to the position of the target in the step S301 to obtain a corresponding phase signal phi (t);
s303, performing phase unwrapping on the phase signal phi (t) obtained in the step S302, wherein the specific steps are as follows: each time the phase difference between successive values is greater/less thanSubtracting 2 pi from the phase to perform phase unwrapping, and recording the phase information obtained after phase unwrapping as phi z (t)；
S304, for phi obtained in step S303 z (t) performing noise reduction and filtering operations, and recording the obtained signal as phi j (t)；
S305, performing the operations of the steps S301 to S304 at intervals of a detection period T to obtain a series of phase dataAs a phase signal of the intermediate frequency signal.
Specifically, in step S4, training the deep neural network specifically includes:
and collecting valve core and valve rod vibration signals of an electromagnetic directional valve in the valve as vibration signal data sets, dividing the vibration signal data sets into training sets, verification sets and test sets after preprocessing, training the deep neural network by using the training sets, verifying the deep neural network by using the verification sets, testing the deep neural network by using the test sets, and storing the predicted optimal deep neural network in test results.
Further, under the working conditions that the rotation speeds are 500rpm, 800rpm, 1000rpm, 1500rpm and 2000rpm respectively, vibration signals of six valve failure parts with the effective abrasion diameters of 20 mu m, 30 mu m, 40 mu m, 50 mu m, 60 mu m and 70 mu m, and vibration signals of six valve failure parts with the normal valve and only valve rod being strained and the degree of strain being very small, tiny, small, medium, serious and very serious are respectively collected and used as a vibration signal data set.
Specifically, in step S4, the deep neural network sequentially includes a one-dimensional convolutional layer, a batch normalization layer, a maximum pooling layer, four residual blocks, an average pooling layer, a full connection layer and a Softmax function, where each set of residual blocks includes 3, 4, 6 and 3 identical network structures, and the Softmax function is used to classify the output result.
The invention also provides a valve fault state sensing and diagnosing system based on terahertz radar, which comprises:
the transmitting module is used for transmitting linear frequency modulation continuous waves to the valve which is working and the surrounding environment and objects of the valve by using the terahertz radar;
the reflection module is used for receiving echo signals reflected by the valve, the surrounding environment of the valve and objects by using the terahertz radar;
the frequency mixing module is used for mixing the echo signal received by the reflection module with the frequency modulation continuous wave transmitted by the transmission module by the terahertz radar to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal;
the network module is used for inputting the phase signals of the intermediate frequency signals obtained by the frequency mixing module into a trained deep neural network;
and the diagnosis module is used for completing the valve core and valve rod fault diagnosis of the electric reversing valve in the valve by utilizing the fault position and the fault degree output by the deep neural network of the network module.
Compared with the prior art, the invention has at least the following beneficial effects:
according to the valve fault state sensing diagnosis method based on the terahertz radar, a mathematical model is provided for providing a theoretical basis for extracting signal vibration frequency from a transmitting signal of the radar, an echo signal after the radar meets a target and a mixed signal of the transmitting signal and the echo signal. Then further explaining the specific steps of extracting the phase signal from the intermediate frequency signal, and the theoretical basis by which the phase signal can characterize the vibration signal. The intermediate frequency phase signal containing the target vibration characteristic is used as an input signal of the neural network, and the neural network is used for diagnosing different fault types of the valve.
Furthermore, the linear frequency modulation continuous wave can be arranged to effectively detect the distance, speed, relative displacement change condition and the like of the target, and has the advantages of high distance resolution, simple structure and high sensitivity. The frequency of the signal is obviously visible to be related to the time t after deriving the signal because of the chirped continuous wave signal. A mathematical model of the transmitted signal is given and the rationality of the patent is verified by formula derivation and simulation analysis of the mathematical model.
Further, by increasing the time delay t based on the transmitted signal d The echo signal can be directly and effectively represented. The echo signals are represented by a mathematical model, and theoretical basis is provided for the derivation of the next intermediate frequency signals, the extraction of vibration frequency and the like.
Furthermore, the frequency of the signal can be greatly reduced by mixing, and the sampling pressure is reduced. The center frequency of the signal before mixing is f c The center frequency of the mixed signal is f b The former is of the order of 10 11 The latter of the order of magnitude is 10 5 About, it can be seen that the sampling rate required for system sampling is greatly reduced.
Furthermore, through a series of algorithms, an intermediate frequency phase signal for removing clutter interference is obtained from the intermediate frequency signal, the intermediate frequency signal in a long-time continuous time period is converted into intermediate frequency phase signals in a plurality of time periods, the intermediate frequency phase signals are used as an input end of the neural network, the processed signal features are more obvious, and the processing of the neural network is easier.
Further, deep learning is based on a data-driven algorithm, and in order to ensure accuracy of deep network prediction, a large amount of relevant data is required to train the network. Further, the data set is divided into a training set, a validation set and a test set: training set data is used for training a depth network, and parameters of a model are obtained through the training of the data; the verification set data are used for selecting super parameters of the depth network and adjusting the model, so that the depth network is ensured to be kept in an ideal learning state during training; the test set is used for evaluating the trained deep network model, the part of data is not available in the training process, and the data is used after the training process is finished. Typically, a plurality of sets of deep network models are trained, and the network parameters with the best predicted results are collected as the final results through testing.
Further, the deep neural network has different applicability for different problems, and in the invention, the neural network is used for processing the working condition displacement signal of the valve core obtained by the terahertz radar and intelligently judging the running state of the workpiece. In order to ensure the accuracy of neural network prediction, a great amount of data aiming at the problems of the invention are required to be collected to train the neural network, so that valve core vibration signals under different working conditions and different wear degrees are collected as data sets to train the network specially.
Furthermore, the convolution operation core in the convolution layer is weight parameter sharing, so that the number of model parameters can be reduced, and meanwhile, the characteristic information in an input vibration signal can be effectively extracted for subsequent operation; the batch normalization layer normalizes the batch mean value-variance of the batch training data currently input into the layer, can effectively accelerate the training and convergence speed of the network, control gradient explosion and prevent gradient disappearance, and meanwhile, the batch normalization considers the distribution of other sample data in the batch, so that the batch normalization can avoid the overfitting of the neural network to a certain extent; the maximum pooling layer plays roles in reducing dimension, removing redundant information, reducing calculated amount and the like, and data of the vibration signal after being preliminarily extracted by the convolution layer can be further screened; each residual error module consists of a plurality of groups of convolution and batch normalization layers and a jump structure, is the core content of the ResNet network, and can effectively solve the problem of network performance degradation caused by deepening of the neural network through the serial connection of 16 residual error modules, and simultaneously solves the gradient problem to improve the network performance; the Softmax function maps the output of the neural network to a value in the range of (0, 1), i.e. outputs the probability that the vibration signal belongs to a certain fault type.
In summary, the invention can accurately measure the weak vibration of the internal components of the valve; all-weather detection can be realized, and automatic diagnosis of valve faults is realized; the measurement efficiency is high and the ultrahigh accuracy can be maintained; the fault position and the fault degree can be accurately judged.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
FIG. 1 is a flow chart of the operation of the present invention;
FIG. 2 is a diagram of an apparatus of the present invention;
FIG. 3 is a deep neural network used in the model of the present invention.
Wherein, 1, the valve; 2. a first terahertz radar; 3. a second terahertz radar; 4. a computer; 5. an electric reversing valve.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it will be understood that the terms "comprises" and "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
Various structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the accompanying drawings. The figures are not drawn to scale, wherein certain details are exaggerated for clarity of presentation and may have been omitted. The shapes of the various regions, layers and their relative sizes, positional relationships shown in the drawings are merely exemplary, may in practice deviate due to manufacturing tolerances or technical limitations, and one skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions as actually required.
The invention provides a valve fault state sensing diagnosis method based on terahertz radar, which comprises the steps of firstly, transmitting linear frequency modulation continuous waves to a valve under operation by the terahertz radar; then, the terahertz radar receives echo signals reflected by the valve to be detected and surrounding objects; mixing the echo signal and the transmitting signal to obtain an intermediate frequency signal, preprocessing the obtained intermediate frequency signal by a computer, extracting a phase vibration signal, and performing noise reduction filtering treatment to reduce the influence of signals reflected by surrounding environment and objects; the preprocessed signals are input into a deep learning neural network trained in advance, so that the defect that the traditional method cannot distinguish fault categories is overcome, and the fault occurrence degree can be effectively and accurately identified; and finally, outputting a fault diagnosis result. The neural network model is an intelligent algorithm integrating feature extraction and pattern recognition, has strong learning capacity, can automatically extract features, is faster and more accurate, can realize automatic detection, has high detection efficiency, can keep ultrahigh accuracy, and can accurately identify the failed part and the failed degree.
Referring to fig. 2, the valve fault state sensing and diagnosing device based on the terahertz radar of the present invention includes a valve 1, a first terahertz radar 2, a second terahertz radar 3 and a computer 4, wherein an electric reversing valve 5 is arranged in the valve 1, the first terahertz radar 2 is used for transmitting linear frequency modulation continuous waves to the valve 1, and the second terahertz radar 3 receives echo signals reflected by the valve 1; the echo signal is then mixed with the transmit signal to obtain an intermediate frequency signal, which is finally fed into the computer 4 for further processing.
Referring to fig. 1, the valve fault state sensing and diagnosing method based on terahertz radar of the invention comprises the following steps:
s1, a first terahertz radar 2 emits linear frequency modulation continuous waves to a valve 1 which is in operation and surrounding environment and objects;
detecting a target by using a Frequency Modulated Continuous Wave (FMCW) radar of a terahertz wave band; the electric reversing valve 5 is a working part in the valve 1, and a valve core valve rod in the electric reversing valve 5 is easy to break down; the signals emitted by the radar are:
wherein a is 0 For the amplitude of the transmitted signal, j is the sign of the imaginary part, f c The range of the center frequency of radar emission is terahertz wave band, T is time variable, B is bandwidth of radar, and T is period of linear frequency modulation.
S2, the terahertz radar 3 receives echo signals reflected by the valve 1 to be tested and surrounding objects;
the echo signals reflected by the radar receiving object are as follows:
where k is the attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary part, f c Is the center frequency of radar emission, t is a time variable, t d For the time delay of the received signal, B is the bandwidth of the radar and T is the period of the chirp.
Time delay t of received signal d The method comprises the following steps:
where R is the distance of the target from the radar and c is the speed of light in vacuum.
S3, mixing the echo signal and the transmitting signal by the terahertz radar 3 to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal b (t) to obtain a phase signal of a plurality of sections of intermediate frequency signals
After mixing and filtering, the radar receives object signals from the R range as follows:
wherein conj is conjugate operation, s (t) is linear frequency modulation continuous wave emitted by terahertz radar to the valve which is working, the surrounding environment of the valve and objects, r (t) is echo signal, k is attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary sign, B is the bandwidth of the radar, R is the distance of the target from the radar, c is the speed of light in vacuum, T is the period of the chirp, T is the time variable, lambda is the wavelength, f b Is the frequency of the intermediate frequency signal b (t), phi b Is the phase of the intermediate frequency signal b (t).
The intermediate frequency signal is a complex signal divided into a real part and an imaginary part.
I.e. for a single object, the intermediate frequency signal b (t) has a frequency f b The phase is phi b Wherein:
wherein B is the bandwidth of the radar, r is the distance between the target and the radar, c is the speed of light in vacuum, T is the period of linear frequency modulation, and lambda is the wavelength.
The signal processing of the intermediate frequency signal b (t) is specifically:
s301, performing fast Fourier transform on the intermediate frequency signal obtained in the step S3, and determining the position of the target according to the peak position obtained by the fast Fourier transform;
according to the formula:
the deformation is carried out to obtain the following components:
wherein f s Is the sampling frequency, n samples Is the number of sampling points, n bin The position of the wave peak is the reflection of the target, and the position of the target is obtained;
s302, carrying out phase solving operation on intermediate frequency signal data belonging to the position of the target in the step S301 to obtain a corresponding phase signal phi (t);
s303, performing phase unwrapping on the phase signal phi (t) obtained in the step S302, wherein the specific steps are as follows: subtracting 2 pi from the phase to perform phase unwrapping every time the phase difference between successive values is greater/less than + -pi, and recording the phase information obtained after phase unwrapping as phi z (t)；
S304, for phi obtained in step S303 z (t) performing noise reduction, filtering and other operations to reduce the influence of surrounding interference noise, and recording the obtained signal as phi j (t)；
S305, performing the operations of the steps S301 to S304 every detection period T to obtain a series of phase dataThe signal is an intermediate frequency phase signal.
S4, inputting the preprocessed signals into a depth neural network trained in advance, wherein the depth neural network is illustrated by taking Resnet as an example, but the depth neural network is not limited to the network model, and can be a LeNet network or an Alexnet network;
the terahertz radar is required to be used for acquiring operation data of valve stems of a certain number of electromagnetic directional valves in advance to train the model so as to be used for subsequent real-time monitoring;
the terahertz radar is used for collecting valve core and valve rod vibration signals which are easy to cause faults of the electromagnetic directional valve, and vibration signals of a normal valve and six valve fault parts which are only worn by the valve core and have effective wear diameters of 20 mu m, 30 mu m, 40 mu m, 50 mu m, 60 mu m and 70 mu m and the vibration signals of the normal valve and six valve fault parts which are only worn by the valve rod and have very small, tiny, small, medium, serious and very serious pull degrees are respectively collected as data sets under five working conditions that the rotating speeds are respectively 500rpm, 800rpm, 1000rpm, 1500rpm and 2000 rpm.
Preprocessing the acquired vibration signals, dividing the vibration signals into a training set, a verification set and a test set, training and testing the deep neural network, storing predicted optimal deep neural network parameters in test results, and taking the optimal deep neural network parameters as a pre-trained deep neural network for subsequent use.
The preprocessing is distance dimension Fourier transformation, phase extraction, phase unwrapping, median filtering and wavelet noise reduction.
Referring to fig. 3, signals collected for different types of fault components described in step S1 are input into a deep neural network to be trained, and feature extraction is performed on the deep neural network. The ResNet network for one-dimensional signals is firstly a one-dimensional convolution layer, a Batch Normalization layer and a Max-pooling layer; four residual blocks follow, each group of residual blocks containing 3, 4, 6 and 3 identical network structures, respectively; following the residual block is an Average pulling layer, an FC layer, and Softmax function.
The activation function is located after the BatchNormalization layer, not specifically described, and by default the ReLU activation function is used.
The Batch Normalization layer is a batch normalization layer, which can prevent gradient disappearance and gradient explosion and accelerate convergence rate; the Max-pooling layer is a maximum pooling layer; the Average pooling layer is an Average pooling layer; the FC layer is a full connection layer; the Softmax function is used to classify the output results.
S5, outputting a fault diagnosis result, wherein the fault diagnosis result comprises a fault part and a fault degree.
In still another embodiment of the present invention, a valve fault state sensing and diagnosing system based on a terahertz radar is provided, where the system can be used to implement the valve fault state sensing and diagnosing method based on a terahertz radar, and specifically, the valve fault state sensing and diagnosing system based on a terahertz radar includes a sending module, a reflecting module, a mixing module, a network module and a diagnosing module.
The transmission module transmits linear frequency modulation continuous waves to the valve which is working, the surrounding environment of the valve and objects by using the terahertz radar;
the reflection module is used for receiving echo signals reflected by the valve, the surrounding environment of the valve and objects by using the terahertz radar;
the frequency mixing module is used for mixing the echo signal received by the reflection module with the frequency modulation continuous wave transmitted by the transmission module by the terahertz radar to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal;
the network module is used for inputting the phase signals of the intermediate frequency signals obtained by the frequency mixing module into a trained deep neural network;
and the diagnosis module is used for completing the valve core and valve rod fault diagnosis of the electric reversing valve in the valve by utilizing the fault position and the fault degree output by the deep neural network of the network module. In summary, the valve fault state sensing diagnosis method and system based on the terahertz radar provided by the invention aim at the problems that the existing vibration measuring method is complex in process, strict in working environment requirement, incapable of realizing all-weather detection, low in measurement accuracy, low in measurement efficiency and the like, and has the following advantages:
(1) The terahertz radar has small volume, particularly the aperture of an antenna can be greatly reduced, the integration can be conveniently realized, the small antenna aperture can obtain narrow wave beams, the directivity is good, the spatial resolution is extremely high, and the test operation is simple;
(2) The high distance resolution capability can be achieved, terahertz is sensitive to micro Doppler characteristics, finer relative displacement changes can be measured, characteristic analysis of micro targets is facilitated, and measurement is more accurate;
(3) The test environment is not strict, the capability of penetrating smoke and dust is stronger, and the test environment is not influenced by extreme weather, so that the test environment can work around the clock;
(4) The effects of positioning and vibration monitoring can be achieved on targets with smaller sizes;
(5) The valve with the medium temperature being too high, the temperature can influence the precision of the traditional measuring method, and the terahertz radar detection adopts non-contact detection, and the temperature can not interfere terahertz echo, so that the method can be suitable for the fault diagnosis of valves with different medium types.
The terahertz ultra-high distance resolution capability is utilized to be sensitive to micro Doppler characteristics, so that finer distance change can be measured, and Doppler characteristic analysis can be performed on the micro target; the device is not affected by extreme weather, and can work around the clock; the volume is small, and the integration is high; providing extremely narrow antenna beam, obtaining advantages of higher antenna gain, better angle resolution and the like, and measuring weak vibration of internal components of the valve. Therefore, the invention realizes non-contact detection by using the terahertz radar, can realize all-weather detection under extreme conditions, has high and accurate detection efficiency and simple process.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above is only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by this, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the claims of the present invention.
Claims (10)
1. A valve fault state sensing diagnosis method based on terahertz radar is characterized by comprising the following steps:
s1, transmitting linear frequency modulation continuous waves to a valve which is working by using a terahertz radar and an object around the valve;
s2, receiving echo signals reflected by the valve and the surrounding environment and objects of the valve by using a terahertz radar;
s3, mixing the echo signal received in the step S2 with the frequency modulation continuous wave transmitted in the step S1 by using the terahertz radar to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal;
s4, inputting the phase signal of the intermediate frequency signal obtained in the step S3 into a trained deep neural network;
and S5, outputting a fault part and the fault degree by using the deep neural network in the step S4, and completing the fault diagnosis of the valve core and valve rod of the electric reversing valve in the valve.
2. The terahertz radar-based valve fault state sensing and diagnosis method according to claim 1, wherein in step S1, the transmitted linear frequency modulated continuous wave S (t) is specifically:
wherein a is 0 For the amplitude of the transmitted signal, j is the imaginary part, f c The range of the center frequency of radar emission is terahertz wave band, T is time variable, B is bandwidth of radar, and T is period of linear frequency modulation.
3. The terahertz radar-based valve fault state sensing and diagnosis method according to claim 1, wherein in step S2, the echo signal r (t) is specifically:
where k is the attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary part, f c Is the center frequency of radar emission, t is a time variable, t d For the time delay of the received signal, B is the bandwidth of the radar and T is the period of the chirp.
4. The terahertz radar-based valve fault state sensing and diagnosis method according to claim 1, wherein in step S3, the intermediate frequency signal b (t) within the range of the target distance terahertz radar distance R is:
wherein conj is conjugate operation, s (t) is linear frequency modulation continuous wave emitted by terahertz radar to the valve which is working, the surrounding environment of the valve and objects, r (t) is echo signal, k is attenuation coefficient of the received signal, a 0 For the amplitude of the transmitted signal, j is the imaginary part, B is the bandwidth of the radar, R is the distance of the target from the radar, c is the speed of light in vacuum, T is the period of the chirp, T is the time variable, lambda is the wavelength, f b Is the frequency of the intermediate frequency signal b (t), phi b Is the phase of the intermediate frequency signal b (t).
5. The terahertz radar-based valve fault state sensing and diagnosis method as claimed in claim 4, wherein the frequency f of the intermediate frequency signal b (t) b And phase phi b The method comprises the following steps of:
6. the terahertz radar-based valve fault state sensing and diagnosing method according to claim 1, wherein in step S3, the signal processing on the intermediate frequency signal to obtain the phase signal is specifically:
s301, performing fast Fourier transform on the intermediate frequency signal, and determining the position of the target according to the peak position obtained by the fast Fourier transform;
s302, carrying out phase solving operation on intermediate frequency signal data belonging to the position of the target in the step S301 to obtain a corresponding phase signal phi (t);
s303, performing phase unwrapping on the phase signal phi (t) obtained in the step S302, wherein the specific steps are as follows: subtracting 2 pi from the phase to perform phase unwrapping every time the phase difference between successive values is greater/less than + -pi, and recording the phase information obtained after phase unwrapping as phi z (t)；
S304, for phi obtained in step S303 z (t) performing noise reduction and filtering operations, and recording the obtained signal as phi j (t)；
S305, performing the operations of the steps S301 to S304 at intervals of a detection period T to obtain a series of phase dataAs a phase signal of the intermediate frequency signal.
7. The terahertz radar-based valve fault state sensing and diagnosing method according to claim 1, wherein in step S4, training the deep neural network is specifically:
and collecting valve core and valve rod vibration signals of an electromagnetic directional valve in the valve as vibration signal data sets, dividing the vibration signal data sets into training sets, verification sets and test sets after preprocessing, training the deep neural network by using the training sets, verifying the deep neural network by using the verification sets, testing the deep neural network by using the test sets, and storing the predicted optimal deep neural network in test results.
8. The terahertz radar-based valve failure state sensing and diagnosing method according to claim 7, wherein vibration signals of a normal valve and six valve failure pieces with effective wear diameters of 20 μm, 30 μm, 40 μm, 50 μm, 60 μm, 70 μm and vibration signals of a normal valve and six valve failure pieces with pull-out degrees of very small, tiny, small, medium, serious, and very serious are collected as vibration signal data sets under the working conditions that rotational speeds are 500rpm, 800rpm, 1000rpm, 1500rpm, 2000rpm, respectively.
9. The terahertz radar-based valve fault state sensing and diagnosis method according to claim 1, wherein in step S4, the deep neural network sequentially comprises a one-dimensional convolution layer, a batch normalization layer, a maximum pooling layer, four residual blocks, an average pooling layer, a full connection layer and Softmax functions, wherein each group of residual blocks respectively comprises 3, 4, 6 and 3 identical network structures, and the Softmax functions are used for classifying output results.
10. A terahertz radar-based valve fault state sensing and diagnosing system, comprising:
the transmitting module is used for transmitting linear frequency modulation continuous waves to the valve which is working and the surrounding environment and objects of the valve by using the terahertz radar;
the reflection module is used for receiving echo signals reflected by the valve, the surrounding environment of the valve and objects by using the terahertz radar;
the frequency mixing module is used for mixing the echo signal received by the reflection module with the frequency modulation continuous wave transmitted by the transmission module by the terahertz radar to obtain an intermediate frequency signal, and performing signal processing on the intermediate frequency signal to obtain a phase signal of the intermediate frequency signal;
the network module is used for inputting the phase signals of the intermediate frequency signals obtained by the frequency mixing module into a trained deep neural network;
and the diagnosis module is used for completing the valve core and valve rod fault diagnosis of the electric reversing valve in the valve by utilizing the fault position and the fault degree output by the deep neural network of the network module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210018701.2A CN114323635B (en) | 2022-01-08 | 2022-01-08 | Valve fault state sensing diagnosis method and system based on terahertz radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210018701.2A CN114323635B (en) | 2022-01-08 | 2022-01-08 | Valve fault state sensing diagnosis method and system based on terahertz radar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114323635A CN114323635A (en) | 2022-04-12 |
| CN114323635B true CN114323635B (en) | 2023-08-01 |
Family
ID=81024371
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210018701.2A Active CN114323635B (en) | 2022-01-08 | 2022-01-08 | Valve fault state sensing diagnosis method and system based on terahertz radar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114323635B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130046883A (en) * | 2011-10-28 | 2013-05-08 | 한국전력공사 | Predictive diagnostic method and system on mechanical integrity of generator stator |
| CN108830127A (en) * | 2018-03-22 | 2018-11-16 | 南京航空航天大学 | A kind of rotating machinery fault feature intelligent diagnostic method based on depth convolutional neural networks structure |
| CN109556503A (en) * | 2018-10-30 | 2019-04-02 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | THz clock synchronization frequency modulation continuous wave nondestructive thickness measuring detection system |
| WO2020244134A1 (en) * | 2019-06-05 | 2020-12-10 | 华南理工大学 | Multi-task feature sharing neural network-based intelligent fault diagnosis method |
| CN112710465A (en) * | 2021-01-04 | 2021-04-27 | 南京航空航天大学 | Wind turbine blade fault classification method based on radar echo features and random forest |
| WO2021217364A1 (en) * | 2020-04-27 | 2021-11-04 | 西门子股份公司 | Fault diagnosis method and apparatus therefor |
-
2022
- 2022-01-08 CN CN202210018701.2A patent/CN114323635B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130046883A (en) * | 2011-10-28 | 2013-05-08 | 한국전력공사 | Predictive diagnostic method and system on mechanical integrity of generator stator |
| CN108830127A (en) * | 2018-03-22 | 2018-11-16 | 南京航空航天大学 | A kind of rotating machinery fault feature intelligent diagnostic method based on depth convolutional neural networks structure |
| CN109556503A (en) * | 2018-10-30 | 2019-04-02 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | THz clock synchronization frequency modulation continuous wave nondestructive thickness measuring detection system |
| WO2020244134A1 (en) * | 2019-06-05 | 2020-12-10 | 华南理工大学 | Multi-task feature sharing neural network-based intelligent fault diagnosis method |
| WO2021217364A1 (en) * | 2020-04-27 | 2021-11-04 | 西门子股份公司 | Fault diagnosis method and apparatus therefor |
| CN112710465A (en) * | 2021-01-04 | 2021-04-27 | 南京航空航天大学 | Wind turbine blade fault classification method based on radar echo features and random forest |
Non-Patent Citations (4)
| Title |
|---|
| A remaining useful life estimation method for solenoid valve based on mmWave radar and auxiliary particle filter technique;Shuo Li, et al;IEICE Electronics Express;第18卷(第20期);1-5 * |
| 基于太赫兹波的轴承防尘盖表面缺陷检测方法研究;潘铁强;黑龙江科技信息;106-107 * |
| 太赫兹雷达目标微动特征提取研究进展;杨琪;邓彬;王宏强;秦玉亮;;雷达学报(01);22-41 * |
| 太赫兹雷达系统总体与信号处理方法研究;李晋;中国博士学位论文全文数据库(07);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114323635A (en) | 2022-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101738611B (en) | Underwater acoustic target signal detection and identification method | |
| CN107678003B (en) | Target detection method under ground wave radar sea clutter background | |
| CN110208806B (en) | Marine radar image rainfall identification method | |
| CN108804740B (en) | Long-distance pipeline pressure monitoring method based on integrated improved ICA-KRR algorithm | |
| CN109190272B (en) | Concrete structure defect detection method based on elastic waves and machine learning | |
| CN102928435A (en) | Aircraft skin damage identification method and device based on image and ultrasound information fusion | |
| CN110082429B (en) | Tunnel lining nondestructive testing auxiliary judgment method combining machine learning | |
| CN107247267B (en) | Angle response correction method based on multi-beam backscattering intensity data and measuring system thereof | |
| CN105891787B (en) | First-order sea clutter detection method based on least square approximation | |
| CN113313146A (en) | Intelligent monitoring method for abnormal operation state of equipment | |
| CN106708009A (en) | Ship dynamic positioning measurement system multiple-fault diagnosis method based on support vector machine clustering | |
| CN111695461A (en) | Radar clutter intelligent classification method based on image frequency characteristics | |
| CN114323635B (en) | Valve fault state sensing diagnosis method and system based on terahertz radar | |
| CN114739669B (en) | Rolling bearing state monitoring method and device based on terahertz radar | |
| CN111694053A (en) | First arrival picking method and device | |
| Fang et al. | Reciprocating compressors intelligent fault diagnosis under multiple operating conditions based on adaptive variable scale morphological filter | |
| CN112801033A (en) | AlexNet network-based construction disturbance and leakage identification method along long oil and gas pipeline | |
| CN108428012A (en) | A kind of fan noise prediction technique of optimization neural network | |
| CN116953087A (en) | Intelligent detection method for bridge segment assembly construction quality | |
| CN111652063A (en) | Gray level stacking chart identification method of ground penetrating radar | |
| CN115932773A (en) | Target angle detection method, device, equipment and medium based on spectrum shape characteristics | |
| Li et al. | Acoustic signal analysis for gear fault diagnosis using a uniform circular microphone array | |
| CN113837163B (en) | Tunnel monitoring method and system based on three-dimensional ground penetrating radar and storage medium | |
| CN117630895B (en) | Submarine forward-looking sonar fault monitoring system based on data analysis | |
| CN117331125B (en) | Split beam data processing method and device, electronic equipment and medium |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Ke Zhiwu Inventor after: Wang Zhongxing Inventor after: Zheng Wei Inventor after: Lin Yuansheng Inventor after: Zhao Zhenxing Inventor after: Zhou Hongkuan Inventor after: Chen Chaoxu Inventor after: Zhang Liuyang Inventor after: Wang Rong Inventor after: Tian Fengshuo Inventor after: Lian Guanghui Inventor after: Wang Xingyu Inventor after: Hao Huibo Inventor before: Ke Zhiwu Inventor before: Wang Zhongxing Inventor before: Zheng Wei Inventor before: Lin Yuansheng Inventor before: Zhao Zhenxing Inventor before: Zhou Hongkuan Inventor before: Chen Chaoxu Inventor before: Zhang Liuyang Inventor before: Wang Rong Inventor before: Tian Fengshuo Inventor before: Lian Guanghui Inventor before: Wang Xingyu Inventor before: Hao Huibo |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |