46 results about "Mixed spectrum" patented technology

According to the invention, the multi-mode time-frequency domain conversion and the deep learning technology are combined, and a multi-component mixed Raman spectrum unmixing method is developed, so that clinical in-vivo and in-situ detection and disease diagnosis of novel Raman probes, instruments and the like are facilitated. The method comprises the following steps: (1) converting a mixed Raman spectrum from a time domain to a frequency domain by using fast Fourier transform (FFT), discrete cosine transform (DCT) and discrete sine transform (DST), and extracting frequency domain features; (2) extracting multi-scale local time-frequency domain characteristics of the mixed spectrum by using short-time Fourier transform (STFT) and discrete wavelet transform (DWT); (3) carrying out spectral unmixing calculation in each mode in combination with a one-dimensional attention mechanism U-shaped neural network model; and (4) fusing various modal unmixing results by using a meta-learning method, and analyzing the weight of each modal to obtain an accurate unmixing spectrum. Compared with a traditional Raman spectrum analysis method, the Raman spectrum multi-component signal unmixing method based on multi-modal time-frequency domain transformation and deep learning can accurately separate independent Raman signals of different tissue structures and biochemical components in a complex environment in a living body, so that the unmixing accuracy of the Raman spectrum multi-component signal is improved. Therefore, convenience is provided for subsequent disease mechanism analysis and diagnosis. The method provides an innovative and potential solution for in-vivo and in-situ detection analysis and disease diagnosis of medical clinical Raman spectroscopy.

The invention relates to the technical field of livestock breeding environment monitoring, regulation and control, and discloses a livestock breeding environment gas dynamic monitoring system. The system comprises a data acquisition module for acquiring data by using a multi-source gas sensor; the gas component analysis module is used for analyzing the data based on a mixed spectrum decomposition algorithm to generate a gas component characteristic matrix; the environment regulation and control decision module inputs the matrix into a multi-mode collaborative decision model to generate a regulation and control instruction; the optimization model construction module is used for constructing a multi-constraint dynamic optimization model to optimize ventilation equipment parameters; and the hierarchical control execution module is used for executing the regulation and control instruction through a hierarchical architecture. The system can comprehensively monitor breeding environment gas, accurately regulate and control ventilation equipment, achieve gas concentration equalization and energy consumption minimization, improve the breeding environment quality, reduce the breeding cost and push intelligent development of the livestock breeding industry.

The invention discloses a full-color LED screen optical adjustment and intelligent energy-saving detection method, and relates to the field of LED screen adjustment and energy saving. Ambient light data is collected through a bionic distributed sensor network, and features such as illumination intensity and color temperature are extracted through PCA dimension reduction; constructing a human eye visual perception model based on the CIE chromaticity diagram, and introducing a visual adaptation factor to dynamically adjust color gamut mapping; dividing a screen dimming area, and calculating an optimal brightness value by adopting a local contrast enhancement algorithm; the RGB ratio is optimized through a genetic algorithm, and the color difference between the mixed spectrum and the target color gamut is minimized; power consumption is monitored in real time, an energy consumption prediction model is established, and an optimization strategy is triggered when a predicted value exceeds a threshold value. The color gamut coverage rate is high, the local contrast ratio is excellent, and the comprehensive energy-saving rate is greatly improved; blue light harm is reduced, screen flicker is eliminated, the visual comfort is improved, the functions of peep prevention and block chain energy consumption evidence storage are achieved, and the display quality, the energy efficiency performance and the use safety of the LED screen are comprehensively improved.

In a method for training a neural network for detecting an object, geometric dimensions of a test object from an object class are captured, and during a time period, recordings of the test object are generated by a plurality of cameras. From the captured geometric dimensions and the generated recordings, occupancy maps are generated. By a radar device, a radar signal is transmitted, and a radar signal reflected by the test object is received. The transmitted radar signal and the received radar signal are mixed into a complex baseband to form a mixed signal. A complex four-dimensional mixed spectrum of the mixed signal is calculated. From the complex four-dimensional mixed spectrum, a first complex two-dimensional partial spectrum and a second complex two-dimensional partial spectrum are calculated. The occupancy maps and the partial spectra are fusioned to form training data. The training data are fed to the neural network.

The invention relates to a water environment pollution source reasoning and tracing method, system and equipment and a medium. The method comprises the following steps: constructing a pollution source characteristic database containing spectrum fingerprints and spatio-temporal information; performing time sequence enhancement on the mixed spectral signal of the monitoring point to extract stable features; a hydrological model is coupled to dynamically simulate pollutant transport paths and probabilities, and a potential contribution source set with weights is generated; analyzing the mixed signal by adopting a deep unmixing network embedded with space-time constraint, and separating and quantifying known source contribution and unknown pollution components; and finally, generating a visual pollution contribution degree spatial distribution diagram and a quantitative traceability report through geographic information mapping. The method solves the problems that complex mixed signals are difficult to accurately analyze, multi-source contribution is difficult to dynamically quantify and unknown pollution components are difficult to effectively recognize in the prior art, and intelligent and accurate traceability of municipal water environment non-point source pollution is achieved.

The application provides a fault identification method and system based on multi-element spectral mineral anomaly analysis, relates to the fields of geology, spectroscopy, petrology and engineering geology, and a specific scheme comprises the following steps: in-situ testing of tunnel surrounding rock is carried out by using a visible light-near infrared, mid-infrared and laser Raman spectrometer, and mixed spectral variation characteristics are extracted; based on the mixed spectral variation characteristics, the mineral composition of the tunnel surrounding rock is identified according to the spectral identification standard of the abnormal mineral combination of the fault zone, quantitative inversion of the minerals is carried out, and the mineral content is obtained; according to the mineral anomaly mode of the fault zone, mineral anomaly analysis is carried out on the mineral composition, and the mineral anomaly type is obtained; based on the differential change of the mineral content and the mineral anomaly type, mineralogical anomaly characteristic analysis is carried out, and the fault identification result is output; the application replaces the experience knowledge of traditional geologists with quantitative information of multi-element spectral mineral anomaly, and further improves the accuracy and efficiency of fault identification in tunnels and underground engineering.

Embodiments may provide photobioreactor system such as a vertical growth enhancing mixing spectrum photobioreactor (1) including but not limited to at least one downcomer tower (22) having spiral impressions (211), an upcomer tower (21), an algae-fluid input (10), an algae downward fluid flow (25) in a downcomer tower; a gas input (18) to drive a system flow (250), a gas upward flow (20) in a upcomer tower; a carbon dioxide gas input (27) having a lower mass flow of carbon dioxide than a mass flow in an upcomer tower, a carbon dioxide gas upward flow (26) in a downcomer tower; a counterflow (230) created with the algae downward fluid flow and carbon dioxide gas upward flow; and perhaps even a collection tank (64) for dewatered algae. Each photobioreactor system may be connected to provide a plurality of photobioreactor systems of which can be automatically controlled.

The application discloses a kind of based on near infrared multiband detection abdominal fat distribution analysis device and method, it is related to fat analysis field, the spectrum of the information of subcutaneous adipose tissue (SAT) mainly reflected by close-range probe acquisition, accurate estimation is combined with pre-training model individualized subcutaneous fat profile parameter. Then, these parameters are used to drive an optical forward propagation model, dynamically predict and generate the SAT contribution spectrum of the individual at the remote probe. Finally, subtract this accurate prediction SAT contribution signal from the mixed spectrum collected by the remote probe, thereby effectively stripping the strong background noise. The signal-to-noise ratio of the residual spectrum obtained is significantly improved, and the visceral fat index can be accurately analyzed through a special correction model, solving the core technical problem that VAT cannot be measured due to SAT signal interference.

This invention relates to the field of computational optical imaging and provides a hyperspectral computational imaging method, system, and related equipment for multispectral acquisition. The method includes: acquiring filter parameters; constructing a spectral simulation reconstruction network; setting a minimum reconstruction error and using the filter parameters to set variables for the annealing algorithm, constructing a simulated transmittance curve to train the spectral simulation reconstruction network, calculating the error until the annealing algorithm calculates the optimal solution and obtains the optimal parameters; training the spectral reconstruction network based on the optimal parameters, simultaneously selecting corresponding filters to form spectral sampling units, and sampling the target spectrum to obtain a mixed spectral image; registering the mixed spectral image to obtain a registered spectral image; and inputting the registered spectral image into the spectral reconstruction network to obtain a reconstructed spectral image. This invention improves the spectral acquisition speed, solves the registration problem of snapshot-type spectral acquisition, and further improves the accuracy of hyperspectral computational imaging.

The application relates to the field of LED health lighting, and discloses a method for LED mixed generation of sunlight-like white light and an LED lamp set. The method comprises the following steps: determining the color points of each LED lamp bead around a blackbody radiation line, independently packaging each LED lamp bead, and comprising LED chips with specific peak wavelengths and various fluorescent powders; determining l target sunlight-like white light color points on the blackbody radiation line, and calculating l groups of luminous flux parameters; selecting a group of luminous flux parameters from the l groups of luminous flux parameters according to a preset target, and using the group of luminous flux parameters to control the m LED lamp beads to form mixed light, i.e. sunlight-like white light. According to the combination of different intensities of various light sources, the application changes the single waveband excitation of fluorescent powder for emitting white light by the existing full-spectrum white light LED into multi-waveband LED chip mixed dimming, fills the gap of the spectrum of about 480 nm, has a higher color rendering index, and has a more continuous mixed spectrum, and can realize sunlight-like dimming along the blackbody radiation line.

The application discloses a kind of based on regularized non-negative least square method's production method carbon-oxygen ratio calculation method.There is following step: S1, obtains stratum common element and carbon-oxygen ratio logging instrument's inelastic gamma standard spectrum;S2, by carbon-oxygen ratio logging instrument to the stratum of place is measured, and the inelastic gamma mixed energy spectrum obtained by measurement is analyzed;S3, establish by each standard energy interval's inelastic gamma standard spectrum and inelastic gamma mixed spectrum count constitutes the contradictory linear equation group;S4, obtains the production of stratum carbon, oxygen element;S5, using the production of stratum carbon, oxygen element obtained in step S4 does ratio, obtains the production method carbon-oxygen ratio not by energy spectrum count statistics influence.The beneficial effect of the application is: improve the accuracy of production method carbon-oxygen ratio and the sensitivity of carbon-oxygen ratio to oil layer identification.

The invention discloses a polymerization reaction intelligent monitoring regulation and control method and device based on spectral analysis, and relates to the technical field of intelligent monitoring, and the method comprises the steps: constructing a particle size-track comparison feature library based on a prior polymerization reaction knowledge library, and combining track feature vector prior thresholds of different particle size intervals to obtain a particle size-track comparison feature library; decoupling the mixed spectral signal of the current polymerization reaction, and outputting spectral signals and track feature vectors of different particle size intervals; and comparing the track feature vectors with a prior threshold value, performing abnormal deviation analysis, and triggering a corresponding directional regulation and control strategy according to an analysis result. The technical problems that an existing polymerization reaction monitoring method is poor in real-time performance and low in precision, and abnormal fluctuation in the reaction process cannot be effectively handled are solved, and the technical effects that intelligent monitoring and signal decoupling based on spectral analysis are achieved, real-time and accurate regulation and control of the polymerization reaction process are achieved, and therefore the product quality and the production efficiency are improved are achieved.

The application discloses a method and device for intelligent monitoring and regulation of polymerization reaction based on spectral analysis, and relates to the technical field of intelligent monitoring.The method comprises the following steps: constructing a particle size-track contrast feature library based on a prior polymerization reaction knowledge base, combining a track feature vector prior threshold of different particle size intervals, decoupling a mixed spectral signal of a current polymerization reaction, and outputting spectral signals and track feature vectors of different particle size intervals; comparing the track feature vectors with the prior threshold, performing abnormal deviation analysis, and triggering a corresponding directional regulation strategy according to an analysis result.The application solves the technical problems that the existing polymerization reaction monitoring method has poor real-time performance, low precision, and cannot effectively cope with abnormal fluctuations in the reaction process, achieves intelligent monitoring and signal decoupling based on spectral analysis, realizes real-time and accurate regulation of the polymerization reaction process, and thus improves the product quality and production efficiency.

The application provides a living body detection method, device and equipment. The method comprises the following steps: obtaining a source domain image and a natural image; performing frequency domain transformation on the source domain image to obtain a source domain spectrum image, performing frequency domain transformation on the natural image to obtain a natural spectrum image, generating a mixed spectrum image based on the source domain spectrum image and the natural spectrum image, and performing inverse frequency domain transformation on the mixed spectrum image to obtain a perturbed source domain image; adjusting network parameters of an initial living body detection model based on the source domain image and the perturbed source domain image to obtain an adjusted living body detection model, and determining a target living body detection model based on the adjusted living body detection model; wherein the target living body detection model is used for living body detection on a to-be-detected image. Through the technical scheme of the application, it can be verified whether the current operation is a real living body operation, and the accuracy of living body detection is relatively high.

The photoelectric device of multimode demultiplexing spectrometer and spectral reconstruction method relate to mode division (demultiplexing) spectroscopy. The photoelectric device is composed of a plurality of photoelectric detectors in cascade, each of which includes an SOI substrate, a silicon oxide layer, a multimode waveguide with two side branches, a two-dimensional material absorber and a metal electrode. The spectral reconstruction method includes a single-mode spectral reconstruction method, a single-sample multimode reconstruction algorithm and a multi-sample resolution enhancement reconstruction method. Different spectral information is modulated on different waveguide modes, then input into the multimode demultiplexing spectrometer, and a photocurrent is generated on each photoelectric detector. By using the spectral reconstruction method, single-mode spectral reconstruction under single sample, one-time multimode mixed spectral reconstruction and resolution enhancement spectral reconstruction under multi-sample are realized. The multimode demultiplexing spectrometer can simultaneously realize mode demultiplexing function and spectral detection capability, greatly reducing the complexity of the wavelength-division-mode-division hybrid multiplexing system.

The application relates to the technical field of signal processing, and discloses an intelligent anti-interference signal processing method and system for a portable laser wind measurement radar, the method comprising the following steps: constructing a physical-inspired feature network, outputting a feature vector and mapping the feature vector to a low-dimensional physical parameter vector, constructing a composite loss function according to the low-dimensional physical parameter vector; constructing a raindrop particle size distribution according to the low-dimensional physical parameter vector, calculating the backscattering cross section and the terminal falling velocity of each particle, obtaining a Doppler frequency shift according to the terminal falling velocity, and obtaining the spectrum of an interference signal according to the backscattering cross section and the Doppler frequency shift; obtaining a pure wind field signal according to the spectrum of the interference signal and a mixed spectrum, and obtaining a horizontal wind vector according to the pure wind field signal. The application realizes high-precision wind speed inversion.

The invention discloses an aliasing spectrum functional group data structured storage method based on multistage wavelet transformation, and belongs to the technical field of road engineering materials. The method aims at solving the problems of functional group signal aliasing and data storage redundancy existing in existing asphalt infrared spectrum data. Comprising the following steps: preprocessing infrared spectrum original data of the asphalt cement to obtain corrected spectrum data, identifying and separating out a plurality of characteristic functional group regions, and performing intensity value enhancement to obtain the whole of the enhanced characteristic functional group regions; a dynamic cascade decomposition system based on a db8 discrete wavelet basis performs adaptive layer decomposition on the enhanced characteristic functional group region to obtain multi-stage wavelet frequency domain data; performing non-uniform sampling and compression to obtain key feature vectors; and constructing a metadata knowledge graph, and carrying out structured storage on key feature vectors of a plurality of feature functional groups in a streaming parallel computing mode by adopting a relational-document hybrid storage architecture. According to the invention, structured data management and lightweight storage of infrared data are realized.

There is provided an imaging device including a pixel array unit (300) configured by arraying, in a row direction and a column direction, a plurality of pixels (304) of five or more types in which wavelength bands of detectable light are different in stages. The plurality of pixels are arrayed such that, at points having any spatial phases on the pixel array unit, mixed spectral characteristics obtained by mixing spectral characteristics of a predetermined number of pixels around the points are substantially same.

The invention discloses a pear maturity nondestructive testing method and system based on mixed spectrum multiple tasks, and belongs to the crossing field of nondestructive testing and spectrum analysis technologies. The prediction method comprises the following steps: acquiring spectral reflectivity data of a target pear fruit, and performing three-view preprocessing to generate an original spectrum, a first-order derivative spectrum and a standard normal variable standardized spectrum; inputting the processed spectral data into a pre-trained mainstream pear variety global model, and synchronously outputting predicted values of hardness and soluble solids; and finally, obtaining a corresponding specific judgment threshold according to the variety of the target fruit, and comprehensively judging the maturity by comparing a predicted value with the threshold. According to the method, the advantages of traditional machine learning and deep learning are fused, the mainstream pear variety global model with the cross-variety generalization ability is constructed, the variety specificity threshold value is combined, rapid, lossless and accurate prediction of the maturity of the multi-variety pears is achieved, and the method is suitable for multi-scene application such as orchard harvesting, storage grading and market circulation.

The application discloses a LIBS multi-distance mixed spectrum classification method based on a deep CNN and sample weight optimization, and relates to the technical field of spectrum analysis. The application constructs a deep convolutional neural network (CNN) model and uses the model for classification of laser-induced breakdown spectrum (LIBS) data, and is suitable for analyzing mixed LIBS spectrum data collected at multiple different distances. The core innovation of the application is that different sample weights are given to spectrum samples at different distances in the CNN training process. In a conventional model training method, equal weights are given to all training set spectrum samples, while in the application, the weight of each spectrum sample is specially designed according to the absolute distance value of the training set spectrum sample and the distance difference between the training set spectrum sample and the test set spectrum sample, so that the weights of spectrum samples at different distances are optimized, and thus the classification effect of the deep CNN model is improved. The application has the advantages of no distance correction, efficient training and high accuracy, and can effectively classify LIBS multi-distance mixed spectrum.

The application belongs to the technical field of image processing, and provides a mixed color correction model construction method, a color deviation prevention image processing method and device. The mixed color correction model construction method comprises the following steps: obtaining a first spectral reflectance characteristic curve according to the color information and corresponding reflection spectrum data of a printing material; obtaining a second spectral reflectance characteristic curve according to the color gamut range and corresponding reflection spectrum data of a target ink on an unengraved printing material; obtaining a third spectral reflectance characteristic curve through the color value and corresponding reflection spectrum data of a multi-level gray background produced by a laser engraving machine; superimposing the target color ink and the multi-level gray background to obtain mixed color values and mixed spectrum data, and then constructing a mixed color correction model. The method and device realize accurate reverse mapping from the target color to the laser parameters and ink parameters, and avoid the occurrence of significant and nonlinear color deviation of the superimposed color ink.

The present application relates to water quality detection technical field, disclose a kind of multi-source spectral composite pollutant identification method and system based on graph depth learning, method contains following steps: S1, obtain multiple single pollution sample, multiple multi-pollution sample and unknown sample to be predicted;S2, obtain the spectral topology feature of multiple single pollution sample, multiple multi-pollution sample;S3, according to the spectral topology feature of single pollution sample, estimate single pollution source prototype vector;According to the spectral topology feature of multi-pollution sample, estimate multi-pollution source prototype vector.It has beneficial effect: the method introduces graph depth learning into the field of pollutant identification of spectroscopy for the first time, and the complex relationship between nodes can be better handled relative to traditional identification model.In order to estimate the prototype vector of unknown mixed spectrum using the prototype vector of known single pollution spectrum in the mixed pollution label missing scene, prototype contrast learning is introduced, and the mixed pollution spectrum representation is effectively extracted.

The present application relates to the technical fields of power plant fuel particle size online detection and automatic sorting based on infrared spectrum, and provides a power plant fuel particle size online detection and automatic sorting system based on infrared spectrum, which comprises a data acquisition module for obtaining a mixed spectrum flow.A difference stripping module performs a space-time difference operation on the mixed spectrum flow, extracts absorption baseline data, and separates out preliminary decoupled scattering data.A photon verification module calculates a residual evaluation matrix using the preliminary decoupled scattering data, and outputs purified scattering data through closed-loop iterative operation.A morphology inversion module performs inversion calculation on the purified scattering data to obtain real particle size characteristic data.A scheduling module drives a sorting device to perform mechanical separation actions according to the real particle size characteristic data, and drives a pulverizing device to perform speed adjustment actions according to the absorption baseline data.The present application solves the problems of spectrum deep coupling and misjudgment caused by the nonlinear changes of moisture and ash concentration under the mixed combustion conditions of multiple coal types.

The invention discloses a cement gas polluted water turbidity detection system based on spectrum analysis, and relates to the technical field of water quality detection, the system comprises a signal acquisition module, a spectrum analysis module, an ion detection module, a data verification module and a result output module; a Raman dual-spectrum sensor is deployed through a signal acquisition module to acquire a mixed spectrum signal containing a chemical substance spectrum signal and a turbidity interference signal, a spectrum analysis module is combined to perform baseline correction and characteristic peak extraction and matching on the mixed spectrum signal, and the pollution component type is judged based on a standard spectrum parameter library. According to the method, the turbidity interference signal and the chemical substance spectral signal are separated, the pollution component signal is obtained through purification, the interference signal and the target signal are effectively stripped in the process, the accuracy of pollution component type identification is ensured, meanwhile, a pure characteristic signal carrier is provided for subsequent data verification, and pollution component traceability has a clear signal basis.

The present application relates to the technical field of hyperspectral remote sensing image processing and spectral information modeling, and particularly relates to a target scene spectral library construction method of frequency domain structured modeling and spectral unmixing. First, spatial structure modeling is performed on the observation image of the target scene to generate a hyperspectral intermediate representation that maintains spatial structure consistency; then, the hyperspectral intermediate representation is corrected for spectral consistency under the guidance of information constraints to suppress unstable disturbances across spectral channels; second, mixed spectral signals in the target scene are separated through spectral unmixing to extract representative spectral components; finally, the representative spectral components are screened and organized to construct a target scene spectral library and output as spectral priori. The present application realizes automatic generation of target scene spectral priori with physical meaning and stability from low-dimensional observation images, and is suitable for hyperspectral reconstruction, target recognition and interpretation and other remote sensing downstream applications that rely on spectral priori information.

The invention belongs to the technical field of pig sound in pig breeding, and particularly relates to a method for discovering respiratory diseases of group health-preserving pigs under the condition of balanced feed nutrition, and the algorithm comprises the following steps: 1, collecting different state audios of pigs, converting the audio time domain waveform into a Mel spectrogram, and obtaining a group health-preserving pig aliasing audio Mel spectrogram; step 2, inputting the aliasing audio Mel spectrogram of the group health-care pigs into the trained TAVM-UNet model, and separating out a single-pig cough / multi-pig cough aliasing spectrogram; 3, converting the single-pig cough / multi-pig cough aliasing Mel spectrogram into a cough aliasing time domain waveform; and step 4, separating the cough aliasing time domain waveform by using an NMF algorithm, and reconstructing the time domain waveform of each kind of cough sound in the multi-pig cough aliasing sound.

This invention discloses a method and system for improving the stability of a coherent dispersive spectrometer based on cross-cancellation, belonging to the field of spectral imaging technology. It includes acquiring two interference spectra with a phase difference of π in an asymmetric coherent dispersive system; separating the local spectrum and interference spectrum through addition and subtraction; performing Fourier transform on both to obtain the local spectrum and interference spectrum in the frequency domain; determining the optical path difference based on the grating line spread function and signal-to-noise ratio condition; and dividing the interference spectrum into a first component region and a second component region; determining the weights using a phase residual optimization method within the frequency range from zero to the optical path difference; and performing a weighted summation of the spectral components of the local spectrum and the first component region to obtain the mixed spectrum after cross-cancellation. This invention further improves spectral resolution through high-frequency stitching and suppresses ringing effects through equalization processing. This invention significantly improves system stability and spectral resolution without increasing system size and complexity.

The invention provides a spectral data separation method and system, and the method comprises the steps: calculating initial non-vegetation spectral data corresponding to to-be-processed mixed spectrum original data according to an initial vegetation coverage rate estimation value and initial vegetation spectral data; based on the initial vegetation spectrum data, the initial non-vegetation spectrum data and the initial vegetation abundance value, performing fitting to obtain initial mixed spectrum fitting data; based on the dual-objective optimization function, the vegetation spectral feature constraint condition and the initial mixed spectrum fitting data, repeatedly executing an iterative optimization solving process to obtain a spectral data separation result; the dual-objective optimization function is constructed by taking the minimum difference between the mixed spectrum original data and the mixed spectrum fitting data as an objective; the vegetation spectrum feature constraint condition is constructed based on the absorption feature and the reflection feature of the vegetation spectrum in the preset wave band. The spectral data separation operation is simplified, and the automation degree and adaptability of spectral data separation are improved.

The present application relates to the technical field of mine gas monitoring, in particular to a device and method for separating and identifying oil-type gas methane and ethane cross interference, comprising a control module, a laser one, a laser two, an optical path transmission device one, an optical path transmission device two, an optical microcavity, a dichroic mirror, a photodetector and a tuning fork sensor; two different voltage waveform signals are generated; two lasers are driven to generate outgoing laser beams with methane and ethane gas spectral absorption wavelengths; the two laser beams are split by the dichroic mirror, the same frequency resonance light signals reflected by the dichroic mirror are collected, and the light signals transmitted by the dichroic mirror are collected; demodulation processing is performed to obtain mixed spectral data of methane and ethane gas, which is input into a spectral analysis model to output independent spectra of methane and ethane after separation of the mixed spectral data. The spectral average absolute error is used as a similarity evaluation index, which can effectively separate and accurately determine the independent spectra of methane and ethane.