843 results about "Bandpass filtering" patented technology

According to the cooling tower early fault early warning method based on vibration state monitoring, vibration signals and working condition labels of key parts of the cooling tower are synchronously collected through multiple channels, and data quality is improved through preprocessing operation such as band-pass filtering and normalization; time-frequency features are extracted in a multi-scale mode through self-adaptive variational mode decomposition and wavelet packet transformation, signal complexity is quantized through energy entropy, and weak fault detection capacity is enhanced; the obtained features are input into a deep belief network after being subjected to principal component analysis dimensionality reduction, and automatic classification and recognition of the equipment operation state are achieved; dynamic early warning grade adaptation is carried out according to an identification result in combination with a working condition label, the environmental adaptability and stability of early warning are effectively improved, the method further has the functions of early warning sample recording and periodic model iterative optimization, and the fault identification precision and robustness in a complex noise environment are remarkably improved.

The invention relates to the technical field of quantum sensing and ocean detection, and discloses an ocean magnetic field sensor based on an NV color center and an ocean magnetic field measuring method.The sensor comprises a cylindrical pressure-resistant cabin, an optical platform arranged in the pressure-resistant cabin and a main control panel; a laser, a beam shaping unit, a dichroscope, a beam shaping lens group, an objective lens, a diamond NV color center chip, a plano-convex lens I, a band-pass filter and a photoelectric detector are mounted on the optical platform; two microstrip loop antennas are mounted on the outer side of the diamond NV color center chip; an FPGA + ARM dual-core processor, a digital lock-in amplifier and two microwave signal generators are installed on the main control board, and the microwave signal generators are connected with the microstrip loop antenna and achieve electromagnetic coupling with the diamond NV color center chip. The invention has the advantages of high sensitivity, good long-term stability and strong anti-interference capability, and provides a new generation of quantum sensing solution for applications such as ocean geomagnetic monitoring and underwater target identification.

The invention relates to the technical field of power system power distribution network fault detection, in particular to an arc fault diagnosis and analysis method based on an artificial intelligence algorithm, and the method comprises the four steps: synchronous data collection and topological excitation, deployment of terminals at transformer area nodes, injection of characteristic current, and synchronous collection of response waveforms; signal preprocessing: separating a key frequency band through double-digital band-pass filtering, and calculating energy ratio and other enhanced fault features; performing multi-source feature fusion and intelligent diagnosis, constructing a vector containing statistical features, time domain distortion features and topology identification, and inputting a gradient boosting decision tree and deep neural network hybrid model to obtain fault confidence and type; and based on fault positioning and verification of topology, scheduling multi-node cooperative monitoring, and determining a fault point in combination with a topological relation. The method improves the data reliability and diagnosis precision, achieves the precise positioning of a fault, and guarantees the safe operation of a power distribution network.

The invention provides a real-time electroencephalogram phase locking closed-loop stimulation method and system, and relates to the technical field of electroencephalogram signal processing. The method comprises the following steps: firstly, acquiring a cortex electrical activity signal and establishing a baseline statistic, then carrying out causal band-pass filtering on the signal, and carrying out sampling and holding substitution in a stimulation artifact window; setting an amplitude and a duration threshold based on the baseline, and performing phase judgment after conditions are met; estimating an average period through zero crossing and calculating an instantaneous phase; outputting a trigger instruction when the three-point crossing judgment is established; and outputting stimulation under the constraint of charge balance and a safety limit value, setting a minimum prohibition period and link abnormity suppression, and recording a hit rate and a response rate for quality evaluation. According to the method, through causal phase estimation, artifact isolation and a schedulable triggering mechanism, the high hit rate of non-stationary signal phase locking stimulation is achieved, and the safety and effectiveness of a stimulation system are ensured.

The invention relates to the technical field of deep learning, in particular to a dialysis hypotension risk prediction system and method based on deep learning, and the system comprises a multi-point signal collection module, a signal processing module, a multi-scale causal expansion convolutional network prediction module, an osmotic pressure correction module and a dialysis plan automatic adjustment module. The method comprises the following steps: synchronously collecting physiological signals of a patient through a finger tip photoelectric volume sensor and a radial artery piezoelectric sensor array, carrying out band-pass filtering and morphological refinement processing on the signals, extracting features, and calculating pulse wave conduction time and a change rate thereof; inputting the characteristic parameters into a multi-scale causal expansion convolutional network to predict the hypotension risk in dialysis; dynamically correcting the prediction result according to electrolyte monitoring data; based on the corrected risk, a hierarchical response strategy is adopted to automatically adjust dialysis parameters, a multi-scale causal expansion convolutional network is adopted to capture physiological parameter changes on different time scales at the same time, and high-precision prediction of hypotension is achieved.

The invention discloses a background noise body wave extraction method and device based on distributed optical fiber acoustic sensing, and the method comprises the steps: obtaining an original phase difference signal along an optical fiber channel through a distributed optical fiber acoustic sensing system, and converting the original phase difference signal into strain rate data according to system demodulation parameters; performing preprocessing such as noise suppression and frequency energy balance on the strain rate data; calculating an autocorrelation function for the preprocessed signal, and superposing autocorrelation results of a plurality of time segments; performing band-pass filtering on the superposition autocorrelation function in a target frequency band, and extracting a body wave reflection response in background noise; and in order to eliminate the influence of the zero-time pulse in the shallow layer or weak reflection signal, carrying out mean value removal processing on the extraction result along the space direction to obtain a zero-offset body wave reflection response. The method can effectively describe the discontinuity from a superficial sedimentary layer and a fracture to a deep mourhua surface and even a lithosphere, and makes up for the defects of a conventional seismic station method.

The invention discloses an encrypted communication system and method based on cognitive multi-carrier spread spectrum, and belongs to the field of covert communication. The encryption communication system comprises a transmitting device and a receiving device. The transmitting device comprises a data encryption module, a constellation encryption modulation module, a multi-carrier spread spectrum module, a frequency hopping module and a cognitive power distribution module. The receiving device comprises a band-pass filtering module, a de-hopping frequency module, a de-spreading module, a constellation decryption demodulation module, a data decryption module and a coherent combination module. The data encryption module is divided into an interleaving coding unit and a double-rule reversible cellular automaton encryption unit. The cognitive power distribution module continuously scans a wireless electromagnetic environment and carries out real-time spectrum analysis to give power spectrum density results sensed under all frequency hopping frequency sets. According to the invention, continuous detection and dynamic response to the spatial frequency spectrum state can be realized, and the communication concealment, the anti-interference performance and the resource utilization efficiency in a complex electromagnetic environment are remarkably improved by utilizing the frequency spectrum hole.

The invention provides a sensor fusion-based pig feed crushing particle size monitoring method, which comprises the following steps of: synchronously acquiring vibration, acoustic emission and optical signals, carrying out band-pass filtering, normalization, feature extraction and wavelet packet decomposition, and dynamically adjusting the confidence coefficient of each channel in combination with working condition perception and a sensor performance knowledge base so as to monitor the crushing particle size of the pig feed. Weighted fusion and conflict evidence modulation of multi-source signals are achieved, the improved Dempster-Shafer evidence theory is introduced to improve the abnormal granularity recognition accuracy, key parameters of crushing equipment are controlled in a linkage mode through a closed-loop feedback mechanism, high-robustness detection and self-adaptive adjustment of the granularity state are achieved, and the method has the advantages of being high in robustness and high in robustness. And the stability and the automation level of the feed crushing process are improved.

The invention relates to the technical field of image processing and biological signal detection, discloses a facial physiological detection method and system based on signal quality driven ROI selection, and aims to solve the problem of signal degradation of a traditional fixed geometric ROI in a complex scene. The method comprises the following steps: collecting a user face video stream through a camera and preprocessing the user face video stream; detecting a face bounding box and dividing the face bounding box into a plurality of sub-regions; calculating the signal-to-noise ratio, the periodic intensity and the motion artifact interference degree of each sub-region; screening an optimal sub-region according to a weighted fusion formula to generate a dynamic ROI mask; extracting a pure rPPG signal from the dynamic ROI mask coverage area; detrending and band-pass filtering are carried out on the rPPG signals, and physiological parameters such as the heart rate and the blood oxygen saturation degree are extracted. The system comprises a face video acquisition module, a face region positioning and segmentation module, a signal quality evaluation module, a dynamic ROI selection module, an rPPG signal extraction module and a physiological parameter estimation module. According to the technical scheme, signal degradation caused by local shielding, illumination abrupt change or attitude offset can be effectively avoided, the signal-to-noise ratio and the stability of the rPPG signal are remarkably improved, and the universality and the robustness of the method are enhanced.

A device for implementing a non-reciprocal bandpass filter is provided. The device includes a plurality of capacitively-coupled shunt resonators. Each capacitively-coupled shunt resonators includes a plurality of superconducting quantum interference devices (SQUIDs) connected to each other, a flux line inductively coupled to the plurality of SQUIDs, and a capacitor connected in parallel to the plurality of SQUIDs.

The invention relates to the technical field of bag type dust collector filter bag abrasion detection, in particular to a filter bag damage non-contact detection method and system based on acoustic array positioning. The method comprises the following steps: arranging an acoustic sensor array in an air purification chamber of the dust remover, and collecting acoustic signals in real time; after band-pass filtering is carried out on the signals, feature vectors of the signals are extracted and matched with a pre-established acoustic fingerprint database, so that a filter bag damage event is recognized; after identification, based on the time difference of arrival of the signal to at least three sensors in the array, combining the coordinates of the sensors and the sound velocity, and utilizing a TDOA positioning algorithm to calculate the space coordinates of the damaged sound source; and finally, matching the coordinates with a pre-stored filter bag coordinate mapping table, determining the number of a specific damaged filter bag, and performing visual highlight marking on a pattern board plane layout of an application layer terminal. The online, real-time, automatic and accurate positioning of the damaged filter bag is realized, the problem that the prior art can only give an alarm but cannot position is solved, and the operation and maintenance efficiency and safety are improved.

The invention belongs to the field of near-inertia internal wave modeling, and particularly relates to a near-inertia internal wave modeling method based on a reverse echo observation array, which comprises the following steps: arranging a reverse echo measurement device array in an observation area, and acquiring echo signals from the seabed to the sea surface; performing data cleaning and band-pass filtering on the signals, and extracting near-inertia internal wave signals; calculating the depth and the relative vorticity of a mixed layer by combining historical temperature-salinity data; calculating a wind input energy flux by using the wind stress data; calculating a net level energy flux by adopting a boundary integration method; judging whether the boundary energy flux influence is smaller than a threshold value or not, and if not, adjusting array layout; and finally establishing a multivariable linear model containing wind energy input, mixed layer depth, relative vorticity and energy flux. According to the method, the temporal-spatial resolution and prediction precision of near-inertia internal wave modeling are effectively improved, the problems that in a traditional method, the boundary effect is remarkable, and multi-factor comprehensive modeling is insufficient are solved, and the method is suitable for long-term efficient monitoring of a large-range marine environment.

The invention provides a target positioning method and system based on a distributed networking radar system, and belongs to the technical field of radars, and the method comprises the steps: controlling a plurality of radar nodes to transmit stepping linear frequency modulation continuous wave pulse signals, and collecting target echoes; carrying out band-pass filtering, pulse compression and fast time synchronization processing on the echoes, and generating interference-free data through a moving target extraction algorithm; realizing cross-node clock synchronization based on a moving target behavior cognition model and a template matching algorithm; constructing a multi-node distance estimation equation set, generating an accurate distance value through calibration broadband synthesis and FFT, and determining a target three-dimensional coordinate; and generating a motion track and speed prediction through Kalman filtering based on the target position sequence. The system comprises a signal emission and acquisition module, a signal processing module, a clock synchronization module, a distance estimation and coordinate fusion module and a trajectory prediction module. The method supports multi-target tracking, improves positioning precision and time synchronization precision, has few errors, and is compatible with any node deployment form.

The invention provides an on-chip terahertz band-pass filter based on an artificial surface plasmon high-order mode, and the on-chip terahertz band-pass filter comprises a dielectric substrate, a feed microstrip line, a first transition structure, a short-branch artificial surface plasmon transmission region, a second transition structure, and a long-branch artificial surface plasmon transmission region. Feed microstrip lines are arranged at the two ends of the front face of the dielectric substrate respectively, each feed microstrip line is sequentially connected with a first transition structure, a short-branch artificial surface plasmon transmission area and a second transition structure, and a long-branch artificial surface plasmon transmission area is located in the middle of the front face of the dielectric substrate. Two ends of the long branch artificial surface plasmon transmission region are respectively connected with a second transition structure, and the whole back surface of the dielectric substrate is covered with a metal layer as a metal ground; the wide-stop-band microstrip antenna has the advantages of being wide in stop band, easy to machine, small in transmission loss and small in crosstalk, does not need punching operation, and is convenient to machine.

The invention provides a brain-computer interface decoding method and system based on multi-modal signal fusion, and relates to the technical field of brain-computer interfaces. The method comprises the following steps: S1, acquiring a multi-mode signal, unifying a time reference and starting a sliding window detection intention; s2, de-prompting, sequential bias checking and permutation correction, artifact suppression and band-pass filtering are carried out; s3, channel and modal quality is calculated, and a reliability weight and a rejection list are generated and marked; s4, multi-scale coding is carried out, cross-modal alignment is realized through learnable time delay compensation, and a first fusion representation is obtained; s5, performing layering and cross attention fusion under reliability gating so as to align and reconstruct constraint to keep semantic consistency, and obtaining joint semantic representation; and S6, sequence-level decoding is carried out to output semantics / states / instructions. According to the method, bias and noise are suppressed, time delay difference and cross-main-body equipment change are resisted, and decoding accuracy, robustness and practicability are remarkably improved.

Methods and systems for mitigating the effects of defects in a quantum processor are provided. A mitigation system includes a quantum processor comprising a plurality of qubits. The system includes a light emitting source that can be tuned to produce light pulses of different wavelengths. The system includes an array of bandpass filters. Each bandpass filter is aligned with a qubit on the quantum processor and has a unique pass band. The system may include a controller configured to receive a selection of a qubit and to tune the light emitting source to emit a light pulse having a wavelength that falls within a range of a bandpass filter that is aligned with the selected qubit. The light pulse is used to scramble an ensemble of strongly coupled two-level system (TLS) in the processor.

The invention discloses a cable partial discharge signal noise separation mode identification method and system, and particularly relates to the technical field of cable partial discharge detection, in a multi-cable stacking environment, partial discharge signals are synchronously collected through multiple types of sensors such as a high-frequency current sensor, an ultrasonic sensor and an electromagnetic antenna; performing band-pass filtering, wavelet denoising and multi-channel synchronous alignment on the acquired signals to suppress background noise and retain key features of partial discharge pulses; separating the mixed signals by adopting independent component analysis; calculating a decision coefficient based on an inter-channel correlation coefficient, a signal-noise power ratio and an amplitude dynamic range, and judging whether to introduce a denoising method based on deep learning; for the waveform after noise separation, executing an amplitude re-calibration process so as to correct amplitude scaling uncertainty, and extracting multi-dimensional time domain, frequency domain and phase features; and a support vector machine and other machine learning algorithms are combined to realize automatic identification of partial discharge types.

The ultralow frequency sensor comprises a vibration detection unit, a power supply voltage stabilization unit and a signal conditioning unit, the power supply voltage stabilization unit is provided with a low dropout linear voltage regulator U20, a voltage converter U19 and a negative voltage regulator U21, and the low dropout linear voltage regulator U20 stabilizes an external power supply as a positive power supply VCC; the voltage converter U19 is used for converting positive voltage of an external power supply into negative voltage and outputting stable negative power supply VEE through the negative voltage stabilizer; the positive power supply VCC and the negative power supply VEE supply power to the signal conditioning unit; the signal conditioning unit comprises a high common mode rejection differential amplifier, a differential-to-single-ended amplifier, a band-pass filter and a differential output driver, a differential signal output by the vibration detection unit sequentially passes through the high common mode rejection differential amplifier, the differential-to-single-ended amplifier, the band-pass filter and the differential output driver, and then a differential voltage signal is output; the method has the advantages of high stability and reliability and the like.

The invention discloses a synchronous motor modular excitation active vibration suppression method for eccentric vibration, and the method comprises the steps: dividing a rotor excitation winding into 2P independent current control excitation modules in the circumferential direction, and respectively arranging an X-direction displacement probe and a Y-direction displacement probe, obtaining a first displacement signal and a second displacement signal of the rotor mass center relative to the stator and a rotor mechanical angular speed; performing band-pass filtering on the first displacement signal and the second displacement signal, and extracting a first displacement first-order component and a second displacement first-order component; calculating the shortest air gap direction angle and the eccentricity amplitude under the stator coordinate system, calculating the eccentricity, and calculating a first-order eccentricity coefficient; calculating a shortest air gap direction angle under a rotor coordinate system, and calculating an equivalent correction coefficient; and according to the equivalent correction coefficient, the first-order eccentricity coefficient, and the shortest air gap direction angle and the mechanical position angle under the rotor coordinate system, the injection amount of each excitation module is calculated, an excitation current instruction is generated and output to the corresponding excitation module, and active suppression of the unbalanced magnetic pulling force is realized.

The invention relates to the technical field of medical treatment, and provides a portable brain-limb cooperative tremor regulation and control method, which comprises the following steps: collecting motion data of a hand of a user; identifying individualized tremor frequency of the user in real time based on the motion data of the hand of the user; based on the identified individualized tremor frequency, band-pass filtering with the individualized tremor frequency as the center is carried out on the tremor signal after high-pass filtering, and the current tremor phase is detected; based on the individualized tremor frequency and the tremor phase, generating percutaneous introduction mode stimulation parameters and starting stimulation; generating transcranial time domain interference stimulation parameters and starting stimulation at a specific time point after the transcranial introduction mode stimulation is started; the transcranial time domain interference stimulation and the transcutaneous afferent mode stimulation are cooperatively controlled for intervention; the tremor intensity change is monitored in real time, stimulation parameters are dynamically adjusted, and closed-loop control is formed. According to the technical scheme, through the integrated design of real-time sensing, dual-mode cooperation and closed-loop regulation, tremor efficient suppression and portable treatment are achieved.

The invention discloses an electroencephalogram fatigue detection method based on fusion of a graph convolutional network and Transform, and belongs to the technical field of artificial intelligence and electroencephalogram signal analysis. The method comprises the steps that multichannel electroencephalogram signals are collected, and band-pass filtering, power frequency notch, independent component analysis, standardization and other preprocessing are conducted on the signals; constructing an inter-channel graph structure based on a Pearson's correlation coefficient, and extracting spatial features by using a graph convolutional network; inputting the spatial features of the plurality of continuous time windows into Transform to carry out time sequence modeling; and finally, realizing fatigue state recognition through a full-connection network and a Softmax classifier. According to the method, the spatial topological structure and the time dynamic evolution of the EEG signal can be modeled at the same time, the accuracy and the real-time performance of fatigue detection are remarkably improved, and the method has good generalization ability and edge deployment ability and is suitable for traffic driving monitoring, intelligent health and other scenes.

The invention provides a bolt ultrasonic axial force detection method and system based on multi-mode signal cooperative processing. The method comprises the following steps: firstly, carrying out band-pass filtering and envelope extraction on a bolt ultrasonic echo signal, and constructing a two-factor adaptive threshold value based on a noise standard deviation and a peak value proportion to realize stable peak value identification; and then screening is carried out by using a peak energy ratio and a waveform correlation coefficient, and multiple reflections and false peaks are eliminated. And aiming at the selected effective wave crest, when the rising edge energy centroid method is adopted to extract sound, the drift caused by noise and coupling change is reduced. In order to enhance the reliability, a frequency domain phase difference verification mechanism is introduced, frequency domain time delay is obtained by calculating an FFT conjugate phase difference between a current signal and a reference signal, and consistency comparison is carried out on the frequency domain time delay and a main measurement sound time result; and when the deviation exceeds the limit, abnormal processing or fusion output is triggered, so that double verification of the time domain and the frequency domain is realized. And finally, obtaining an axial force value by combining temperature compensation and a calibration model. The method improves the stability and adaptability of ultrasonic axial force measurement.

The invention discloses a rapid power calculation method suitable for an electric energy meter and an acquisition terminal, the electric energy meter and the acquisition terminal, and relates to the technical field of electric power measurement. According to the method, voltage and current signals are sampled, direct-current components are removed, and harmonic interference is suppressed through band-pass filtering; estimating the frequency of the power grid in real time by using zero-cross detection and linear interpolation; processing the filtered voltage and current signals by adopting a delay cancellation model based on a rotation matrix, and eliminating a negative frequency component to generate a corresponding orthogonal vector; and synchronously calculating active power and reactive power through algebraic operation according to the orthogonal vector. According to the method, establishment time delay introduced by a traditional low-pass filter is avoided, rapid tracking of a power value can be achieved when the load power is subjected to step change, electric energy metering deviation in the transient process is effectively reduced, and meanwhile calculation accuracy and robustness under the power grid frequency fluctuation working condition are improved through a frequency self-adaption mechanism.

The invention belongs to the technical field of distributed optical fiber sensing, and relates to a phase-frequency noise suppression optical frequency domain reflection measurement method and system without an auxiliary interferometer. The measurement method comprises the following steps: acquiring positive and negative double-sideband sweep-frequency light with the same sweep-frequency range and opposite sweep-frequency directions; dividing the positive and negative double-sideband sweep frequency light into double-sideband detection light and double-sideband local oscillation light according to a preset proportion; configuring a pseudo-random noise code, and modulating the pseudo-random noise code to the double-sideband probe light to obtain spread spectrum double-sideband probe light; the spread spectrum double-sideband detection light enters the sensing optical fiber and is reflected back to the backward Rayleigh scattering light, and the backward Rayleigh scattering light and the spread spectrum double-sideband local oscillation light interfere with each other to output up-down sweep frequency beat frequency signals; separating the upper and lower sweep-frequency beat-frequency signals to obtain an upper sweep-frequency beat-frequency signal and a lower sweep-frequency beat-frequency signal; and after the upper sweep frequency beat frequency signal and the lower sweep frequency beat frequency signal are respectively decoded, a beat frequency signal is separated through band-pass filtering, the beat frequency signal is subjected to frequency mixing processing to obtain a frequency mixing signal after phase frequency noise suppression, and the frequency spectrum of the frequency mixing signal is analyzed.

The invention relates to the technical field of mechanical measurement and sensing, in particular to a silo eccentric unloading monitoring and deviation rectifying method and system.The method comprises the steps that force sensors are installed below all supporting legs of a silo, original load signals are collected, noise is removed through band-pass filtering, and the real-time loads of the supporting legs are obtained; based on the supporting legs, the silo is divided into sectors, material level data of the sectors are obtained, the weight of the sector of the longitudinal gradient is constructed, and the weighted load is calculated; and constructing an eccentric vector based on the weighted load sequence, and mapping the weighted load into a two-dimensional eccentric vector component by using the angle coordinate of each supporting leg. According to the silo eccentric discharging monitoring and deviation rectifying method provided by the invention, the weighted load is mapped into the two-dimensional eccentric vector component based on the angle coordinates of the supporting legs, linear fusion is carried out by combining the two-dimensional eccentric vector component with the material level vector component, the combined eccentric coefficient is obtained, and the coefficient can quantify the eccentric degree and output the eccentric direction angle; and a clear physical basis is provided for subsequent correction decision.

The invention relates to a fault monitoring and protection technology for a flexible direct current transmission system, in particular to a fault detection and polarity judgment method and system for the flexible direct current transmission system, and the method comprises the steps: firstly constructing a differential current and a zero-mode current through positive electrode current and negative electrode current, and carrying out the preprocessing through band-pass filtering; then, in the short-time detection window and the long-time confirmation window, key features of difference and zero mode are extracted by adopting an adaptive truncation and robust statistical method, and zero mode compensation is realized in combination with covariance correction; and finally, through multi-scale fusion of short window and long window results, a fusion mean value and variance are calculated, a confidence interval is constructed, and rapid fault detection of the flexible direct-current power transmission system and accurate determination of the polarity of the positive electrode and the negative electrode are realized. According to the method, rapidity and robustness are taken into consideration, interference of noise and abnormal points is effectively suppressed, rapid and accurate fault identification of the flexible direct current system can be realized in a complex operation environment, and technical support is provided for operation safety and protection control of a flexible direct current power grid.

The invention provides an electroencephalogram signal denoising and artifact-removing method. The method comprises the following steps: acquiring original electroencephalogram signal data; preprocessing the original electroencephalogram data; performing preliminary denoising on the preprocessed electroencephalogram data through channel-by-channel singular spectrum analysis to obtain a global denoising signal matrix; performing band-pass filtering on the global de-noised signal matrix to obtain a filtered signal matrix; iCA decomposition is carried out on the filtered signal matrix, artifact components are automatically removed, effective components are reserved, a clean electroencephalogram signal is reconstructed based on the reserved effective components, and the denoised signal matrix is output. The method is particularly suitable for the real-time processing requirement of portable few-channel electroencephalogram equipment, the application range is wide, and the signal fidelity is high.

The invention relates to a metal component defect depth detection system and method based on sweep frequency pulsed eddy current. The system comprises a pulse eddy current probe module, a pulse signal excitation module, a current amplification module, a voltage amplification module, a band-pass filtering module and a signal acquisition module. The pulsed eddy current probe module comprises an excitation probe and a receiving probe which are respectively used for transmitting excitation signal pulsed eddy current and receiving sampling data; the pulse signal excitation module is connected with the upper computer module and the current amplification module, and is used for receiving the control command and exciting pulse signal current to control probe detection; the current amplification module is used for amplifying the pulse current; the voltage amplification module is connected with the receiving probe and the band-pass filtering module and is used for amplifying the induced voltage; the band-pass filtering module is used for frequency screening; and the signal acquisition module is responsible for performing format conversion on the sampling data and uploading the sampling data to an upper computer. The system can realize excitation and signal acquisition for different depth areas, and accurately determine the characteristic frequency of defect depth.

The invention belongs to the technical field of ground penetrating radar signal processing, and relates to a ground penetrating radar signal denoising method based on optimal proportion-kurtosis joint masking, which comprises the following steps: carrying out short-time Fourier transform on a single-channel ground penetrating radar echo signal to obtain a complex time-frequency graph and a power spectrum thereof; calculating an optimal proportion masking index; calculating the frequency domain kurtosis of each frame according to the amplitude of the complex time-frequency graph in the time-frequency domain, and adaptively determining a kurtosis threshold; constructing an optimal proportion masking matrix, constructing a kurtosis masking matrix by using the kurtosis threshold, and multiplying the optimal proportion masking matrix and the kurtosis masking matrix point by point to obtain a final masking matrix; the final masking matrix is multiplied by the complex time-frequency graph, then inverse short-time Fourier transform is executed, and a preliminary denoising time-domain signal is reconstructed; and applying zero-phase finite impulse response band-pass filtering to obtain a final high-signal-to-noise-ratio ground penetrating radar signal. The method does not need to additionally collect a reference channel, is low in calculation complexity, and can remarkably improve the signal-to-noise ratio of the ground penetrating radar data and the target identifiability.

The invention discloses an emotion recognition method based on multi-modal signal fusion, and relates to the technical field of emotion recognition, and the method comprises the steps: obtaining an electroencephalogram signal and a peripheral physiological signal to form a multi-modal signal, and carrying out the preprocessing of down-sampling, baseline correction and band-pass filtering on the multi-modal signal; based on the preprocessed signal, extracting a difference entropy feature and a power spectrum density feature; mapping the extracted features into a standardized space grid to generate a feature tensor with a uniform structure; performing adaptive weighting on the mapped feature tensor by using a frequency band fusion attention mechanism, and generating a frequency band weight through global average pooling and a full connection layer; inputting the weighted features into a full connection layer of the parameterized hypermatrix, and performing feature compression and modeling through matrix decomposition and reconstruction; a multi-task learning framework is adopted, classification results of emotion titer and awakening degree are output at the same time based on compressed features, and multi-task collaboration is optimized through a shared feature layer and a dynamic loss weight.