75 results about "Signal envelope" patented technology

The invention provides a high-voltage circuit breaker abnormity identification method based on a vibration signal envelope line. The method comprises the following steps: acquiring an operation vibration signal of a high-voltage circuit breaker, generating vibration data by a conditional GAN, filtering the vibration signal by wavelet transform, extracting the envelope line of the vibration signal and associating an operation state label to form a sample library; an AM-CNN-based high-voltage circuit breaker abnormity identification model is constructed, and sample library data is used for training so as to predict the operation state of the circuit breaker on line; and perfecting the sample library according to the prediction state and updating the circuit breaker abnormity identification model. According to the method, the fault state of the high-voltage circuit breaker is diagnosed by automatically extracting the fault features of the vibration signals, the error state and the unknown fault state are judged through self-learning so as to continuously improve the abnormal recognition precision of the circuit breaker, power transformation operation and maintenance personnel are assisted to make decisions, and the operation state recognition efficiency of the high-voltage circuit breaker is improved.

The invention discloses a self-adaptive parameter control system and method for a halfpace field rice transplanter based on full-hydraulic transmission, particularly relates to the technical field of hydraulic parameter control, and aims to solve the technical problem of dynamic power competition between a walking system and a planting device when the full-hydraulic transmission rice transplanter works in a halfpace field. Working pressure signals of a walking motor and a planting device are monitored in real time, the divergence degree of envelope lines of the two pressure signals is calculated so as to recognize the power competition state of a system, and when power competition is detected, the required power of two execution mechanisms is dynamically calculated, and a hydraulic pump displacement distribution coefficient is generated; and then the interaction characteristics of the outlet pressure of the main pump and the output flow signal and the transmission characteristics of the pressure signal of the executing mechanism are fused, the dynamic balance degree of the system is generated, the displacement distribution coefficient is corrected, and finally the opening degree of the walking motor and the proportional valve of the planting device is synchronously adjusted based on the corrected coefficient. Accurate distribution and dynamic balance control of the power of the hydraulic system are achieved.

The invention discloses a miniature front scattering type visibility meter capable of being carried on an unmanned aerial vehicle, which is applied to the field of visibility meters and is characterized in that modulated light rays are irradiated to a reflector I through an emission light path collimating lens cone and then are reflected out of a detection cabin along the direction of an emission channel, and when passing through atmospheric dispersion substances on the lower side of the detection cabin, the light rays are reflected out of the detection cabin along the direction of the emission channel; part of light rays are scattered, enter a receiving chamber through a receiving channel, are reflected by a second reflector, penetrate through a receiving light path collimating lens barrel and irradiate a receiving detector, received electric signals pass through a signal amplification circuit and then are subjected to AD sampling, analog signals are converted into digital signals, the digital signals are sent to a control chip, the digital signals are filtered, and the digital signals are sent to the control chip. The forward scattering visibility meter has the advantages that signal envelope is extracted through transformation, wavelet transformation and noise reduction are performed, extinction coefficients and visibility are finally calculated, the forward scattering visibility meter is miniaturized through special design of a light path system, the use requirements of small size, small wind resistance and light weight are met, and the problem that the visibility meter is difficult to apply to an unmanned aerial vehicle is solved.

The invention discloses a communication method based on a low-noise high-dynamic full-wave radio frequency channel amplification circuit, and relates to the field of wireless communication and radio frequency integrated circuits. The method comprises the following steps: firstly, sampling a baseband signal and constructing a multi-dimensional time sequence feature vector containing a historical state and a change rate; predicting a signal envelope peak value in a future time window by using a lightweight long short-term memory network (LSTM); a dynamic bias control signal is then generated based on reverse mapping and parasitic parameter compensation. Digital delay is introduced into a radio frequency main path, so that bias voltage is pre-loaded to a grid electrode of an amplifier before a radio frequency signal arrives, and negative delay control is realized; and finally driving the amplifier through asymmetric filtering. According to the invention, the problem of signal truncation distortion caused by response lag in the traditional dynamic bias technology is solved, the amplifier is in a deep AB-class or B-class state of low quiescent current for a long time while ensuring no distortion in a high dynamic range, the power consumption and the thermal noise are greatly reduced, and the signal-to-noise ratio of the system is improved.

The application relates to the technical field of underwater detection, in particular to an underwater vehicle autonomous detection system and method based on seabed seismic wave signals. The system comprises a three-axis seismic wave signal acquisition module, a signal superposition amplification module, a signal filtering module, a signal envelope extraction module, an environmental noise integration module and a threshold comparison determination module. The acquisition module is orthogonally arranged to receive three-axis seismic wave signals and convert the signals into charge signals, the superposition amplification module completes signal conversion and superposition, the filtering module suppresses noise, the envelope extraction module extracts a detection value, the noise integration module obtains a noise integral value, the comparison module compares the two values, and when the detection value is greater than the integral value, an underwater vehicle existing signal is output. The application uses seabed seismic waves as a detection source, has strong anti-interference performance, works in a passive mode, is concealed and has low power consumption, can accurately capture weak seismic wave signals of the underwater vehicle, realizes autonomous detection of the underwater vehicle, and is suitable for long-term unattended deployment on the seabed.

The application discloses a kind of calibration method and device for rotary variation soft decoding, calibration method includes: according to N times oversampling frequency respectively sampling rotary variation feedback sine and cosine feedback signal envelope surface and obtaining positive, cosine oversampling signal;Positive, cosine oversampling signal is handled to obtain the zero position offset of rotary variation positive, cosine envelope signal;Positive, cosine oversampling signal is handled to obtain the amplitude of rotary variation positive, cosine envelope surface;Positive, cosine oversampling signal is handled to obtain the phase deviation of rotary variation positive, cosine envelope surface;Through the amplitude, zero position offset, phase deviation obtained by calculation, the envelope surface of rotary variation positive, cosine is compensated, and the signal after compensation is sent into phase-locked loop to obtain the angle and rotational speed information of rotary variation.The application obtains accurate amplitude, zero position offset, phase difference in turn, and then carries out error compensation, finally obtains corrected positive, cosine signal, effectively guarantees the efficiency of motor.

The application provides a fracture patient muscle mass detection method based on electromyographic signal analysis, and relates to the technical field of medical treatment. The method comprises the following steps: acquiring electromyographic signals of a target muscle and its adjacent muscles of a patient; when the correlation coefficients of the target muscle and the adjacent muscle signals and their envelope lines meet preset conditions, judging whether the target muscle is interfered by the adjacent muscles based on the signal difference degree; when the adjacent muscles interfere, extracting independent frequency component sequences of the target muscle and the adjacent muscle signals, removing interference components based on the time domain correlation coefficients and the frequency overlap degree of each independent frequency component group, and obtaining a real target muscle signal; determining the muscle recovery state of the patient based on the amplitude variation trend of the real target muscle signal sequence of the patient in a target period, and the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment in the signal envelope line of the target sub-sequence of the sequence, and formulating a rehabilitation training scheme accordingly. The application can more accurately determine the muscle recovery state of the patient.

The application discloses a kind of self-adapting nonlinear impairment compensation methods based on low-cost optical performance monitoring, and its technical scheme main points are as follows: in transmitting end, low-speed mark is modulated to the envelope of high-speed data signal on top signal;A section of zero-power gap is inserted into signal, after digital-analog conversion and electro-optical conversion are successively carried out to data, into optical fiber link;In receiving end, the signal received is successively carried out photoelectric conversion, analog-digital conversion, IQ orthogonalization, clock recovery;Combining the step number of FD-Volterra-CDE back propagation process in different optical transmission link state parameter range N span Nonlinear effect introduced by the calculation of nonlinear impairment causes optical signal-to-noise ratio penalty;The adaptive nonlinear impairment compensation method in the application solves the problem of too high calculation complexity in the traditional nonlinear impairment compensation process, and by establishing the LUT of the step number N step , filter number N k In the process of FD-Volterra-CDE back propagation in different optical transmission link state parameter range, the calculation complexity is effectively reduced while the compensation accuracy is guaranteed.

Methods for detecting power system events in an electrical power system. An example method comprises determining a signal envelope for each of at least first and second power system signals and performing density-based spatial clustering of a series of points formed by combining respective values of the signal envelopes for at least the first and second power system signals. The example method further comprises detecting one or more power system events by identifying outlier points or groups of points in the spatial clustering. Some methods may further comprise automatically classifying power system events by collecting a data set comprising, for each detected power system event, power system signal features corresponding to the detected power system event, and classifying each of the detected power system events using the data set and a machine-learning-based classification algorithm, where said classifying comprises determining a classification label from among two or more predetermined classification labels.

The application discloses a kind of modulation signal envelope extraction algorithm and electronic equipment, wherein algorithm includes the following steps: S1, according to modulation data, generate waveform image and obtain the image area corresponding to waveform image;S2, extract the original boundary of image area, obtain boundary pixel coordinate point set;S3, the coordinate point set after screening processing is handled to boundary pixel;S4, based on the coordinate point set after screening processing, complete envelope extraction, obtain the envelope set of modulation signal.The present application converts discrete data into waveform image, extracts target area using image segmentation algorithm, extracts area edge point coordinates using edge extraction algorithm of image processing, and then converts irregular area boundary at pixel level into an ordered coordinate set after screening, reduces data volume and improves the accuracy of subsequent envelope extraction, effectively solves the shortcomings in the existing commonly used algorithm;The present application has the characteristics of fast detection speed, safety and reliability, good practicability, high accuracy and the like.

Bandwidth adaptation in a transmission circuit is provided. The transmission circuit includes a power amplifier circuit that amplifies a radio frequency (RF) signal based on a modulated voltage, an envelope tracking integrated circuit (ETIC) that generates the modulated voltage based on a modulated target voltage, and a transceiver circuit that generates the RF signal and the modulated target voltage. The RF signal may have a wide modulation bandwidth but the ETIC may have a lower bandwidth limit, which can cause distortion in the modulated voltage and the RF signal. In this regard, the transceiver circuit is configured to generate the modulated target voltage in a lower bandwidth than the bandwidth limit when the modulation bandwidth of the RF signal exceeds the bandwidth limit of the ETIC. As such, the transmission circuit can process the RF signal across the wide modulation bandwidth without causing distortion in the RF signal.

Embodiments of the present disclosure provide a signal envelope detection circuit and chip. The signal envelope detection circuit comprises: N cascaded first amplification circuits, an input end of a first-stage first amplification circuit being used for receiving a target input signal; M cascaded attenuation circuits, an input end of a first-stage attenuation circuit being used for receiving the target input signal; M second amplification circuits, an input end of each second amplification circuit being correspondingly connected to an output end of one attenuation circuit; and a plurality of rectifiers, an output end of each first amplification circuit and an output end of each second amplification circuit each being connected to an input end of one rectifier. Output ends of all of the rectifiers are connected together and are used for outputting a target output signal. N and M are both integers greater than 1, and amplification multiples of the N first amplification circuits are the same as amplification multiples of the M second amplification circuits. The embodiments of the present disclosure can expand the dynamic range of the signal envelope detection circuit and improve detection accuracy.

The present application relates to the technical field of aquaculture monitoring, and discloses a Chinese octopus health state monitoring system based on voiceprint recognition, which comprises a multi-source sensing module for acquiring synchronous acoustic signals and environmental parameters, a signal purification module for adjusting a step length to restore residual signals by using a step length control formula based on a feedback delay phase error, an envelope demodulation module for separating breathing rhythm and chewing frequency by using Hilbert transform dimension reduction, a rhythm tracking module for calculating a physiological cycle and feeding back an error to the signal purification module, a dynamic affine module for obtaining a target vertex by remodeling vertices of a final feasible polyhedron by using synchronous dissolved oxygen data and synchronous water temperature data, and a state decision module for mapping a feature vector to a dynamic geometric space to generate a monitoring evaluation result. The present application effectively isolates physiological compensation interference induced by physical field natural fluctuation by implanting an environmental nonlinear compensation mechanism into an acoustic feature judgment space, and reduces the false alarm rate of health evaluation under complex working conditions.

Method and system for automatic modulation classification (AMC) use time-series voltage signals representative of radio frequency (RF) signal envelope and frequency components as input features to a deep learning-based neural network, which enables classification of both modulation type and symbol rate without requiring in-phase and quadrature (IQ) demodulation. A feature extraction circuit captures RF signal envelope amplitude and frequency using stub-based sensing, and a Long Short-Term Memory (LSTM) neural network processes these features in a digitized form to classify the modulation and symbol rate with high accuracy and minimal latency.

This invention belongs to the field of wireless communication and signal processing technology, and relates to a digital predistortion method, system, and application for power amplifiers in shortwave communication systems. This invention can cover the wide dynamic range of the power amplifier's operating range in shortwave communication systems, maintaining excellent linearization compensation across the entire power range. It effectively improves modulation signal distortion, spectral broadening, and adjacent channel interference, significantly reducing the system bit error rate and improving the quality of the transmitted signal, thus meeting the requirements for spectral and linearity specifications. This invention eliminates the need for multiple predistortion models and complex switching logic, greatly reducing system implementation complexity and storage resource overhead. When power levels switch or the signal envelope changes rapidly, it avoids additional distortion introduced by abrupt changes in predistortion parameters, ensuring continuous and stable linearization. Furthermore, this invention supports online adaptive updates, enabling real-time adaptation to changes in the nonlinear characteristics of the power amplifier, improving system robustness and environmental adaptability.

This invention discloses a method and apparatus for heart sound segmentation based on a Hidden Markov Model (HMM) cardiac cycle. The method involves: extracting the signal envelope from the heart sound signal using Hilbert's algorithm; obtaining the peak positions of the first heart sound (S1) and the second heart sound (S2) after Hilbert envelope extraction based on preset baselines for different leads; locating the peaks based on the cardiac cycle; extracting the diastolic phase based on an improved HMM; locating the optimal time span of heart sound signals S1 and S2 using an improved Viterbi algorithm; and segmenting the original heart sound signal using the time span and the peak positions of S1 and S2. This invention employs an improved HMM and an improved Viterbi forward algorithm to calculate the duration of heart sound intervals S1 and S2, combined with the cardiac cycle, to accurately locate the peak positions of heart sound intervals, thus improving the performance of heart sound segmentation.

The application discloses a kind of outer wall waterproof layer thickness detection methods based on ultrasonic, it is related to building engineering nondestructive testing technical field, including using ultrasonic detection equipment to collect the original ultrasonic echo signal of the surface of the outer wall waterproof layer to be measured, baseline correction is carried out to signal to eliminate the system error caused by hardware characteristics, and the preprocessed ultrasonic echo signal is obtained.Initial wave peak position reflected on the surface of waterproof layer is identified, and the effective signal segment containing interface echo information is intercepted.The signal envelope is extracted, and the local peak and trough positions are determined, and multiple candidate echo intervals are divided.The fusion characteristic value of the time domain and frequency domain characteristics of each candidate echo interval is calculated, and the target echo interval corresponding to the main interface of the waterproof layer and the base material is selected.The waterproof layer thickness is obtained by calculating the two-way travel time and combining the ultrasonic propagation velocity.This method can eliminate hardware error interference, accurately locate the effective echo and main interface echo, and improve the accuracy of outer wall waterproof layer thickness detection.

An electronic device may include wireless circuitry having an amplifier configured to receive a radio-frequency signal generated from a baseband signal, a first adjustable load component coupled to an output of the amplifier, a second adjustable load component coupled to the output of the amplifier, and a control signal generator configured to output one or more control signals for tuning the first and second adjustable load components based on an envelope of the baseband signal or the radio-frequency signal. The first adjustable load component can provide a first tuning range covering a first subrange of an instantaneous signal envelope of the baseband signal or the radio-frequency signal, whereas the second adjustable load can provide a second tuning range covering a second subrange of the instantaneous signal envelope of the baseband signal or the radio-frequency signal. The first and second tuning ranges are combined to provide an extended tuning range.

The application discloses an intracranial pressure detection method, which comprises the following steps: acquiring regional signal envelope data corresponding to a preset acquisition region, the acquisition region being set as at least two key brain regions, each key brain region corresponding to one regional signal envelope data; inputting the regional signal envelope data into a corresponding intracranial pressure estimation model which has been pre-trained, and outputting initial intracranial pressure data of each region; extracting an intracranial pressure correlation feature vector based on the initial intracranial pressure data of each region, inputting the intracranial pressure correlation feature vector into a contribution index prediction model which has been pre-trained, and outputting a contribution index corresponding to the initial intracranial pressure data of the region; and calculating target intracranial pressure data based on the initial intracranial pressure data of each region and the contribution index. Thus, the dynamic change difference of intracranial pressure in different brain regions is captured, and the accuracy and reliability of intracranial pressure detection are improved.

The embodiment of the invention relates to the technical field of power systems, in particular to a transformer partial discharge ultrasonic three-dimensional positioning method and device and electronic equipment. The method comprises the following steps: acquiring original data including partial discharge signals acquired by ultrasonic sensors arranged on the outer wall surface of a transformer; performing interference signal filtering on the original data to obtain a preprocessed time domain partial discharge signal array; extracting a signal envelope of the time domain partial discharge signal array; performing time axis inversion operation on the signal envelope to obtain an inversion envelope signal, and superposing the inversion envelope signal according to propagation delay from each position to the sensor to obtain an energy focusing signal; and determining the position corresponding to the signal with the maximum energy in the energy focusing signals as the partial discharge source position. According to the method, high-precision three-dimensional positioning of partial discharge of the transformer can be realized.

A method for encoding envelope information is provided. In some implementations, the method involves determining a first downmixed signal associated with a downmixed channel associated with an audio signal to be encoded. In some implementations, the method involves determining energy levels of the first downmixed signal for a plurality of frequency bands. In some implementations, the method involves determining whether to encode information indicative of the energy levels in a bitstream. In some implementations, the method involves encoding the determined energy levels. In some implementations, the method involves generating an energy control value indicating that energy levels are encoded. In some implementations, the method involves generating the bitstream, wherein the energy control value and the information indicative of the energy levels are usable by the decoder to adjust energy levels associated with the first downmixed signal.

The invention relates to the technical field of measurement and testing, and discloses a pulse ultrasonic sensing device and method for fluid anomaly detection, and the method comprises the steps: a programmable pulse generator in an application-specific integrated circuit generates a short pulse sequence with preset parameters, and the short pulse sequence is amplified by a power amplifier and then is transmitted to a fluid medium by an ultrasonic transducer. In a transmitting gap, the transmitting-receiving switching circuit switches the system to a receiving mode, and echo signals are conditioned through the low-noise preamplifier, the programmable band-pass filter and the automatic gain control circuit in sequence. The conditioned signals are digitalized through an analog-to-digital converter array, pulse shaping, time sequence alignment, signal accumulation and digital filtering operation are carried out in a digital processing module, then signal envelopes are extracted by an envelope detector, feature vectors are calculated, and finally abnormity judgment is carried out by a judgment logic unit. According to the invention, rapid ultrasonic detection and accurate judgment of abnormal signals are effectively realized.

The invention discloses a mobile phone sonar identity authentication method based on multiple breathing states, which comprises the following steps that: an intelligent equipment mobile phone transmits an ultrasonic signal which cannot be heard by human ears to the chest part of a human body, and receives an echo signal of chest movement generated by breathing movement at the same time; carrying out band-pass filtering and abnormal value cleaning treatment on the echo signal to obtain a clean high-frequency chest movement signal; judging the breathing state of the preprocessed thoracic cavity movement signal, extracting a signal envelope by using different parameters, and dividing the signal into one breathing cycle of'inspiration-clearance-expiration 'through a designed breathing cycle division mechanism; on the basis of a periodic signal, extracting a Mel-frequency cepstrum coefficient and first-order and second-order differential characteristics thereof, and a generalized cepstrum coefficient and first-order and second-order differential characteristics thereof in a period; extracting frequency domain transmission characteristics of thoracic cavity displacement; and inputting the two types of features into a double-flow neural network for further feature fusion to realize identity authentication. The method is non-contact, low in cost and good in privacy protection.

The embodiment of the present application provides a method and device for improving the sound quality of a loudspeaker, relates to the technical field of media, and can improve the low-frequency sound effect of the loudspeaker and improve the sound quality of the loudspeaker. The method comprises the following steps: performing frequency division on an input signal of the loudspeaker to obtain a first low-frequency input signal and a first high-frequency input signal, wherein the input signal is a time-domain signal; performing transient detection on the first low-frequency input signal to determine whether the first low-frequency input signal is a transient signal; if the first low-frequency input signal is a transient signal, performing signal envelope modulation on the first low-frequency input signal to obtain a second low-frequency input signal; and then determining an output signal of the loudspeaker according to the second low-frequency input signal and the first high-frequency input signal. The starting voltage of the second low-frequency input signal is greater than the starting voltage of the first low-frequency input signal, and the loudness of the second low-frequency input signal is greater than the loudness of the first low-frequency input signal.

The invention relates to the technical field of bearing detection, in particular to a rolling bearing defect detection method and system based on encoder feedback, and the method comprises the steps: collecting a vibration signal of a to-be-detected rolling bearing, and obtaining a rotating speed signal through a coaxial encoder; selecting a defect impact dominant frequency band for the vibration signal according to the spectral kurtosis and carrying out band-pass filtering, and then carrying out envelope demodulation on a filtered signal to obtain an envelope signal; generating a speed measurement pulse according to the rotating speed signal, determining an equal-angle sampling mark and establishing time-angle mapping; resampling the envelope signal angular domain according to the mark to form an angular domain envelope sequence; order spectrum analysis is carried out on the angular domain envelope sequence to obtain an envelope order spectrum, feature orders of an outer ring, an inner ring and a rolling body are determined according to bearing structure parameters, and harmonic peak values of the feature orders of the outer ring, the inner ring and the rolling body are extracted; and calculating three-class order peak ratio characteristics by using the full-spectrum amplitude and normalization, inputting the three-class order peak ratio characteristics into a multi-core support vector machine model, and outputting defect classes. According to the invention, the problems of difficult identification caused by feature drift of the bearing under a non-stationary condition and strong dependence of the diagnosis process on expert experience can be solved.

This invention discloses an automatic identification method for the initial arrival of vibration events based on multi-module filter integration. The main steps are as follows: First, the continuously recorded raw signal is preprocessed and then segmented. For each segmented signal, the effective frequency distribution of vibration energy is determined. Then, operations are sequentially performed, including embedding an anti-aliasing low-pass filter, downsampling the signal, embedding an adaptive detrending term filter, embedding a median filter, and embedding a borderless delay filter. Next, the signal envelope energy is calculated based on the borderless delay filter. Finally, a discriminant filter is embedded, and the signal is normalized to obtain the final identification feature function. This invention establishes an automatic picking algorithm integrated with multiple filters. While reducing noise, it constructs a feature function, requiring only the most basic signal parameters and effectively reducing noise interference. This improves the accuracy of automatically picking the occurrence of disaster events.

The application relates to a signal reconstruction method based on a MEMS high-precision ultrasonic sensor and related equipment, comprising the following steps: envelope extraction is performed on an original echo sequence of the MEMS high-precision ultrasonic sensor to obtain a signal envelope line, and peak point detection is performed on the signal envelope line to obtain a peak time point sequence; the original echo sequence is segmented and intercepted based on the peak time point sequence to obtain multiple echo slices, waveform symmetry analysis is performed on each echo slice, effective echo slices meeting a symmetry threshold are screened out from the multiple echo slices to obtain a target echo section; point-by-point amplitude compensation is performed on the target echo section based on a preset standard waveform template to generate a compensated waveform, and the compensated waveform is subjected to phase alignment splicing to obtain a reconstructed signal, thereby solving the technical problem of how to improve the recognition accuracy of effective waveform segments in the ultrasonic echo signal reconstruction process.

The invention discloses a broadband digital multi-beam joint detection method and system based on large beam amplitude selection, and belongs to the field of signal processing. Digital multi-beam forming is carried out on AD channel echo data of a receiving array plane; uniformly dividing the broadband signal into N paths of parallel narrowband channels according to a preset frequency interval by adopting a multi-filtering channelization technology; for the same channel, comparing each range gate signal amplitude of all beams on the channel, selecting a value with the maximum amplitude as the amplitude of a virtual detection beam of the channel, and performing signal detection; in each channel, according to the start of the signal envelope, comparing the average amplitude at the initial position of the envelope, determining a wave beam with the maximum amplitude as the arrival direction of the signal, marking the wave beam number of the signal as the wave beam, and switching the original data of the wave beam as the input of fine measurement; and in each channel, measuring each parameter of the signal by using the data of the selected beam.

The present application belongs to the technical field of ultrasonic testing, and provides an ultrasonic phased array defect detection system and method for laser penetration overlap welding, which comprises determining the amplitude threshold value between the layers of the laser penetration overlap welding workpiece; instructing the ultrasonic phased array probe to emit sound waves to the laser penetration overlap welding workpiece to obtain echo signals; extracting the signal envelope line of the preprocessed echo signals through Hilbert transform; determining the super-threshold signal segment with an amplitude greater than the amplitude threshold value in the signal envelope line; extracting the time domain features and frequency domain features of the super-threshold signal segment, and inputting the time domain features and frequency domain features into a pre-trained double-channel CNN model to obtain the probability of incomplete fusion; through the double-channel CNN model, the time domain features and frequency domain features are mined, the time correlation law and frequency distribution mode are learned through parallel processing, and then the time domain and frequency domain features are fused, so that the subtle differences between the incomplete fusion defects and normal interfaces in the time domain and frequency domain features can be amplified.

The application discloses a kind of outer wall waterproof layer thickness detection methods based on ultrasonic, it is related to building engineering nondestructive testing technical field, including using ultrasonic detection equipment to collect the original ultrasonic echo signal of the surface of the outer wall waterproof layer to be measured, baseline correction is carried out to signal to eliminate the system error caused by hardware characteristics, and the preprocessed ultrasonic echo signal is obtained.Initial wave peak position reflected on the surface of waterproof layer is identified, and the effective signal segment containing interface echo information is intercepted.The signal envelope is extracted, and the local peak and trough positions are determined, and multiple candidate echo intervals are divided.The fusion characteristic value of the time domain and frequency domain characteristics of each candidate echo interval is calculated, and the target echo interval corresponding to the main interface of the waterproof layer and the base material is selected.The waterproof layer thickness is obtained by calculating the two-way travel time and combining the ultrasonic propagation velocity.This method can eliminate hardware error interference, accurately locate the effective echo and main interface echo, and improve the accuracy of outer wall waterproof layer thickness detection.