1622 results about "Infrared" patented technology

The invention belongs to the technical field of photovoltaic cables, and particularly relates to a photovoltaic cable extrusion control method. The method comprises the following steps: S1, high-frequency data acquisition: introducing temperature-pressure-speed model data to obtain dynamic extrusion temperature, melt pressure and extrusion speed parameters in an extrusion process; an infrared on-line diameter measuring instrument is introduced to acquire real-time data of the outer diameter of a cable insulating layer and the outer diameter of a sheath layer; establishing an influence intermediate value, wherein the influence intermediate value comprises heat conduction delay time, melt flow delay time and equipment response delay time; establishing a traction speed-melt pressure-temperature-predicted outer diameter quaternary coupling model; according to the method, a quaternary coupling model is established by combining real-time data of the outer diameter of an insulating layer and the outer diameter of a sheath layer on the basis of a speed-temperature-pressure acquisition model, and the problem that repeated vicious cycle adjustment is easily caused by passive logic of'adjustment after waiting for deviation exceeding the standard 'is solved.

The invention discloses an artificial intelligence-based refining process efficiency optimization and real-time temperature control method and system, and the method achieves the efficient optimization and precise temperature control of a refining process through the innovation of four steps: 1, breaking through the limitation of traditional single-point monitoring, and deploying a multi-modal sensing array and a feed port near infrared spectrum sensor according to three-dimensional coverage; process-raw material two-dimensional data are collected in real time and associated, and the problem of data islands is solved; 2, innovatively constructing a'process-raw material 'two-dimensional feature vector, predicting a finished product rate by adopting an XGBoost model, and judging a dynamic power adjustment demand in combination with a threshold value to replace a traditional experience decision; thirdly, on the basis of the value output by the DQN model, an optimal dynamic temperature curve is calculated in a staged mode, and then the problem of fixed curve stiffness is solved through a generated precise power instruction; and fourthly, a three-level aging standard reaching mechanism is created for the first time, and an effect closed loop is formed corresponding to a differentiated scheme of model strengthening, local compensation and cross-furnace repairing.

The invention discloses a multi-mode fruit sugar degree nondestructive testing method, device and system and a medium, and the method comprises the steps: preprocessing an input fruit near infrared spectrum image and a fruit visible light image to obtain a one-dimensional near infrared spectrum input image tensor and a two-dimensional visible light input image tensor; a pre-trained multi-modal deep learning fusion model is utilized to obtain the predicted fruit sugar degree, and the multi-modal deep learning fusion model is pre-trained to establish an input near infrared spectrum input image tensor and a visible light input image tensor. The method comprises a spectral feature extraction network, an image feature extraction network, a multi-modal feature fusion network and a regression prediction network. The objective of the invention is to solve the problem of limited prediction precision caused by the fact that single spectral information is susceptible to noise, illumination conditions, peel thickness, water content and other factors, and improve the accuracy of fruit sugar degree nondestructive testing.

The invention relates to the technical field of pot body temperature monitoring, and discloses a pot body surface temperature real-time monitoring system based on an infrared sensing technology. The system comprises a temperature data acquisition module, a feature extraction module, a temperature safety analysis module, a safety temperature range construction module, a deviation analysis module, a communication adjustment module and a temperature uniformity analysis module. The temperature data acquisition module combines the temperature data acquired by the infrared sensor and preset environment parameters to generate a dynamic temperature distribution model; the feature extraction module processes the model to obtain a compression temperature feature flow; a temperature safety analysis module generates a temperature control decision priority atlas according to the temperature control decision priority atlas; the safe temperature range building module builds a safe temperature corridor according to the atlas and generates a control strategy; the deviation analysis module analyzes the temperature deviation to generate abnormal mark data; the communication adjustment module dynamically allocates bandwidth transmission data; and the temperature uniformity analysis module detects temperature uniformity fluctuation and judges whether optimization control is performed or not. The system realizes precise monitoring and intelligent control of the temperature of the pot body.

The invention discloses an embedded fiber grating sensor self-calibration strain monitoring system and method.The system comprises a fiber grating sensor, an optical fiber demodulator and a stress buffering type self-calibration mechanism, and the fiber grating sensor is embedded into a propeller blade prefabricated mounting groove through a metal embedded part and a gradient modulus adhesive layer; the plurality of fiber grating sensors are symmetrically distributed in a propeller blade prefabricated mounting groove in pairs, and infrared rays of each pair of symmetrical fiber grating sensors are converged at the circle center of a propeller hub to form a dynamic reference point; the optical fiber demodulator is integrated on the propeller hub and is connected with the plurality of optical fiber grating sensors; the stress buffering type self-calibration mechanism integrates a double-shaft flexible hinge and a shape memory alloy, and interface stress buffering and fiber grating sensor position collaborative self-calibration are achieved. The interface stress is effectively reduced through the gradient modulus glue layer, and the shape memory alloy drives the double-shaft flexible hinge to efficiently correct the offset of the fiber grating sensor in real time and buffer the vibration.

The invention relates to the technical field of petrochemical engineering, in particular to a method and equipment for predicting the yield of crude oil cut distillate oil, a medium and a product. According to the method, near infrared spectrum data of to-be-predicted crude oil is obtained, a corresponding key spectrum variable is determined, the key spectrum variable is input into a target prediction model corresponding to the type of the to-be-predicted crude oil, and the cut distillate yield output by the model is obtained. Thus, interference of other parameters on prediction accuracy is reduced through the key spectral variables with high correlation degree with the cut distillate yield, and the exclusive prediction model is constructed for different types of crude oil, so that interference of heterogeneity between crude oil types on prediction accuracy is reduced, the adaptability of the prediction model to different crude oil characteristic changes is improved, and the prediction accuracy is improved. Therefore, the prediction model can rapidly and accurately predict the yield of the cut distillate oil of the crude oil, and the prediction accuracy and prediction efficiency of the yield of the cut distillate oil are improved.

The invention relates to the technical field of Brassica rapa polysaccharide detection, and discloses a quantitative detection method of Brassica rapa polysaccharide based on near infrared spectroscopy. The method comprises the following steps: acquiring near infrared spectrum data of a brassica rapa sample, and dividing detection stages by combining spectral characteristics to obtain a plurality of detection stage identifiers; performing differential mapping on the spectrum load spectrum according to the identifier to obtain an optical parameter set of each stage; constructing a staged near infrared spectrum quantitative model by using the set; performing importance weighting on the original spectral data according to the identifier to generate optimized spectral data; and extracting polysaccharide related characteristics from the optimized spectral data through a staged model, and outputting a predicted value of the content of the Brassica rapa polysaccharide. According to the method, through stage division, differential mapping and data weighting, the spectral data utilization rate and the model suitability are improved, a targeted technical path is provided for quantitative detection of Brassica rapa polysaccharide, and the method can be used for quality control of Brassica rapa related products.

The invention provides a brain function evaluation method and system based on fNIRS and eye movement feature fusion, and the method comprises the steps: testing a to-be-tested person under a pre-constructed social scene normal form, and obtaining test data which comprises functional near infrared spectrum data and eye movement data; extracting time dynamic characteristics of brain function activation and brain function network space organization mode characteristics from the functional near infrared spectrum data, and constructing a near infrared brain function characteristic set; for a near-infrared brain function feature set corresponding to the functional near-infrared spectrum data and an eye movement space-time feature set corresponding to the eye movement data, constructing features in each near-infrared brain function feature set and features in the eye movement space-time feature set into feature pairs, and calculating a correlation coefficient of each feature pair, feature fusion and screening are realized based on correlation coefficients; and inputting the screened features into a classification evaluation model to obtain the severity of the behavioral dysfunction and a brain function analysis value.

An inspection system detects defects within a component using eddy current thermography. An infrared camera detects heat from a defect within the component generated by the eddy current interfering with the defect. The eddy current is generated in the component by a magnetic field applied to the component by an inductor assembly. The inductor assembly includes a magnetic core having magnetic core material and two arms, each arm including a coil to generate the magnetic field. The inductor assembly is configured to apply the magnetic field at an angle offset from a horizontal axis of the magnetic core such that the eddy current flows at an angle within the component relative to the offset angle. Further, the magnetic core material can be shaped to apply the magnetic field at the offset angle or to accommodate complex component shapes.

The invention provides a near-infrared model cross-device application method based on multi-source fusion optimization, and aims to solve the problem that various neural network models constructed by a high-precision near-infrared spectrometer cannot be applied to portable devices when being migrated to the portable devices. And the model performance is reduced due to differences of an optical system structure, a wavelength range, a signal-to-noise ratio, an instrument response characteristic and the like. According to the method, the near infrared spectrum data and the conventional quality parameters are subjected to feature level fusion, and the cross-equipment spectrum deviation is calibrated and compensated by using the stability of the conventional quality parameters, so that efficient knowledge migration from high-precision equipment to portable equipment is realized in the fused feature space. The problem of cross-device feature dimension mismatching is solved through a stable feature importance evaluation-based wave band alignment method; a small sample optimization strategy and a hyper-parameter automatic search technology are combined, so that the transfer learning effect is improved, the performance of a portable equipment detection model is remarkably enhanced, and meanwhile, the modeling cost is greatly reduced.

The invention discloses double-sided coated glass. The double-sided coated glass comprises a glass substrate, a first coating layer and a second coating layer, the first coating layer is arranged on the front surface of the glass substrate and sequentially comprises an anti-reflection layer, an ultraviolet reflection layer and a middle and far infrared reflection layer from the glass substrate to the outside; the second coating layer is arranged on the back surface of the glass substrate and sequentially comprises a near-infrared reflecting layer and a near-infrared regulation and control layer from the glass substrate to the outside, and the near-infrared regulation and control layer is used for regulating and controlling the reflectivity of the near-infrared reflecting layer. According to the design of the front anti-reflection layer of the double-sided coated glass, through continuous transition of the particle size and the refractive index of nanoparticles, the visible light reflectivity is remarkably reduced, and the light transmittance of the glass is kept to be 85% or above and is superior to that of existing composite coated glass; meanwhile, through the synergistic effect of the ultraviolet reflecting layer, the middle and far infrared reflecting layer and the back near-infrared reflecting layer, the ultraviolet blocking rate can be larger than or equal to 99%, the comprehensive infrared blocking rate can be larger than or equal to 90%, and the technical problem that light transmission and blocking performance of traditional glass are difficult to consider at the same time is solved.

The invention provides a bottle sorting and data monitoring method and system and bottle recycling equipment. The method comprises the following steps: collecting a visible light image and near infrared spectrum data of a bottle target, and preprocessing the collected visible light image and near infrared spectrum data; extracting a fusion spectrum feature vector of the pre-processed visible light image and the near infrared spectrum data, and generating a feature matrix for material classification; inputting the feature matrix into a lightweight convolutional neural network model deployed in an edge calculation unit, and obtaining a recognition result of the material and form of the bottle; a sorting control instruction is generated based on the recognition result, and motion space state representation is constructed in combination with current state information of the mechanical arm; and inputting the spatial state representation into a path optimization model based on deep reinforcement learning, and outputting and executing an optimal sorting path and a coordination action sequence. According to the method, the precision and robustness of bottle identification can be improved, the recycling efficiency of renewable resources can be improved, and the recycling cost can be reduced.

The invention discloses a binocular thermal imaging image fusion method based on dynamic compensation, and relates to the technical field of infrared imaging, a dynamic compensation operation module is constructed, a pixel displacement matrix of a right eye image is generated according to a real-time position deviation value, real-time geometric correction is performed on the right eye image before the image is output, and the image fusion precision is improved. Through a real-time dynamic registration and progressive fusion mechanism, the problem of ghosting in binocular thermal imaging observation is effectively solved, a thermal radiation feature extraction and cross-channel synchronous analysis technology is adopted, the feature recognition bottleneck of a traditional visible light registration method in a thermal imaging scene is broken through, and the real-time dynamic registration and cross-channel synchronous analysis technology is achieved. In combination with a dynamic offset compensation and temperature adaptive progressive fusion algorithm, sub-pixel-level parallax correction is realized while the integrity of original thermodynamic data is maintained, and continuous ghosting interference caused by binocular image space deviation during human eye observation is eliminated. And real-time observation requirements of targets with different distances can be met without modifying an optical hardware structure.

The invention relates to the technical field of cashmere blended yarn preparation, and discloses near infrared ray absorption and emission fiber and cashmere blended yarn equipment which comprises a rack, a pre-carding unit and a main carding unit are arranged in the rack, and the pre-carding unit is communicated with the main carding unit through an air duct pipeline. The pre-carding unit and the main carding unit each comprise a cylinder and a working roller, a plurality of open grooves are formed in the side wall of the rack, sliding sleeves are installed in the open grooves, first wedge-shaped blocks are installed at the ends of the working rollers, an arc-shaped frame is installed on the rack, an electromagnetic mechanism is installed on the inner wall of the arc-shaped frame, and an iron shaft body is installed in the arc-shaped frame. According to the technology and equipment for absorbing and emitting the fiber and cashmere blended yarn through the near infrared rays, through transmission of the electromagnetic mechanism, the iron shaft body, the second wedge-shaped block and the first wedge-shaped block at the end of the working roller, online, dynamic and micro-adjustment of the carding gauge between the working roller and the cylinder is achieved; therefore, the non-uniformity of the fiber layer can be automatically adapted.

The invention discloses a method for identifying insect pests of tobacco leaves. The method comprises the following steps: S1, acquiring near infrared spectrum data and visible light image data of the tobacco leaves; s2, pre-processing parameters are dynamically adjusted according to the data quality indexes, smoothing processing and derivative calculation are carried out on the near infrared spectrum data, and spectrum feature vectors are obtained; s3, processing the visible light image data through a feature extraction algorithm to generate an image feature vector; s4, splicing the preprocessed spectral feature vector and the preprocessed image feature vector to form a fused feature vector; s5, a convolutional neural network model is used to process the fusion feature vector, and the convolutional neural network model comprises an attention mechanism so as to identify the insect pest type; and S6, outputting a pest recognition result. The method can solve the problems that in the prior art, the morphological characteristics of the tobacco diseases cannot be comprehensively captured due to data singleness, near infrared spectrum data are easily interfered by environmental factors, and the characteristics are lost after preprocessing, so that the accuracy of identifying the insect pests of the tobacco is improved.

The invention belongs to the technical field of boiler heating surface high-temperature corrosion monitoring, and relates to a boiler heating surface high-temperature corrosion online monitoring method and system based on data fusion. The method comprises the following steps: acquiring an electrochemical signal, infrared thermal image data and acoustic emission characteristics of a key corrosion sensitive area of a boiler heating surface; based on data collected by a sensor network, a corrosion rate dynamic prediction model composed of a physical driving sub-model and a data driving sub-model is adopted to obtain a final corrosion rate prediction value; fusing the infrared thermal image data and the acoustic emission features, calculating a regional corrosion risk confidence coefficient according to the fused infrared thermal image data and acoustic emission features, and outputting a confidence coefficient level according to the regional corrosion risk confidence coefficient; and outputting an early warning signal according to the final corrosion rate predicted value and the confidence level, and dynamically adjusting combustion parameters. Accurate prediction of high-temperature corrosion of the boiler heating surface is realized, and corrosion development of the boiler heating surface is actively inhibited.

The invention discloses a near infrared spectrum chemical analysis model cross-equipment construction method based on feature decoupling, and relates to the technical field of near infrared spectrum analys.The method comprises the steps that near infrared spectrum data of different bauxite samples are obtained, and a near infrared spectrum training sample set is constructed; sequentially cascading the feature extractor, the feature decoupling module and the label predictor to construct a network architecture of a near-infrared spectrum chemical analysis model, and training by taking the near-infrared spectrum data as input and the chemical content label as output to obtain the near-infrared spectrum chemical analysis model; and detecting the near infrared spectrum data by using the trained near infrared spectrum chemical analysis model to obtain the chemical content of the bauxite sample to be detected. According to the method, stable features which are only related to chemical components and are not influenced by equipment interference can be decoupled and extracted from near infrared spectrum data, and real robust migration of the model between different equipment is realized.

A pre-stressed pultruded glass fiber reinforced thermoplastic lumber suitable for demanding construction environments, particularly marine and outdoor applications, features a solid core composed of a thermoplastic, such as PVC, blended with 15% to 40% by weight of pelletized chopped strand glass fibers. Pre-tensioned elongated continuous fiber reinforcements are embedded within and coextensive with this core via pultrusion. The presence of pelletized and fragmented glass fibers within the core enhances the mechanical coupling between the core and the continuous reinforcements, facilitating the transfer of tensile stress to the core as compressive pre-stress. An outer co-extruded cap stock provides protection against UV and IR radiation and enhances impact resistance. The resulting lumber exhibits a high modulus of elasticity (exceeding 800,000 psi), at least 60% fiber length retention, improved durability, and resistance to common wood degradation factors without the use of toxic chemicals.

The invention provides a traditional Chinese medicine key medicinal component content detection method and device, and is applied to the technical field of component detection. The method comprises the following steps: acquiring near infrared spectrum data of a plant traditional Chinese medicine sample to be detected, wherein the near infrared spectrum data is used for indicating molecular vibration information; inputting the near infrared spectrum data into a traditional Chinese medicine key effective component quantitative correction model to obtain the target component content of the plant traditional Chinese medicine sample to be detected; wherein the quantitative correction model for the key medicinal components of the traditional Chinese medicine comprises a feature extraction module and a content prediction module, the feature extraction module is used for extracting spectral features of the near infrared spectral data, and the content prediction module is used for predicting the component content of an input sample based on the spectral features. And the content prediction module comprises a nonlinear mapping relation between spectral characteristics and component contents.

The invention relates to the technical field of optical filters, and provides an infrared ray gradient optical filter and a preparation method thereof, and the preparation method comprises the following steps: providing a substrate; coating the upper surface of the substrate for the first time to form a first reflecting layer; the upper surface of the first reflecting layer is coated for the second time to form a spacer layer, the cross section of the spacer layer is of a wedge-shaped structure, and the thickness of the wedge-shaped structure is linearly increased in the preset direction; and carrying out third coating on the upper surface of the spacing layer to form a second reflecting layer, thereby obtaining the infrared linear gradual change optical filter. The surface, away from the substrate, of the second reflecting layer is used for receiving incident light, and the surface, away from the first reflecting layer, of the substrate is used for emitting transmission light. And the second reflecting layer is formed by alternately laminating first refractive index material film layers and second refractive index material film layers. The infrared ray linear gradual change optical filter can continuously and dynamically adjust the spectral resolution, and the wavelength application range is wide.

The present invention discloses a laminated insulating glass with local high infrared transmittance and a vehicle containing the same. The laminated insulating glass includes an outer glass plate, an inner glass plate and a thermoplastic interlayer; the outer glass plate has a first surface and a second surface, and the inner glass plate has a third surface and a fourth surface; the thermoplastic interlayer is sandwiched between the second surface and the third surface; the laminated insulating glass also includes an infrared transmittance-enhancing layer and a heat-insulating layer, and the infrared transmittance-enhancing layer and the heat-insulating layer are not on the same glass surface; the heat-insulating layer includes a film removal area. The present invention meets the use requirement that the attenuation of the laser radar or infrared camera signal passing through the laminated insulating glass is no more than 3dB. While improving the accuracy of the laser radar or infrared camera, it can significantly reduce the energy consumption of air conditioning and improve the comfort of the driver and passengers.

The present invention discloses a device and method for testing the flow corrosiveness and thermal stability of molten salt, which belongs to the field of thermal property testing of energy storage high-temperature molten salt. The device includes: an experimental test subsystem, a heating constant temperature subsystem, and an information acquisition and control subsystem; the experimental test subsystem includes a W-shaped quartz tube, an air pump, and a molten salt tank; the W-shaped quartz tube includes a vertical first quartz tube, a second quartz tube, and a third quartz tube, and the lower ends of the three vertical quartz tubes are connected to accommodate the high-temperature molten salt to be tested. The present invention detects the dynamic and static thermal stability of the molten salt by detecting the penetration rate of infrared rays through the capillary tube, and can also detect the corrosiveness of the molten salt in two states of long-term immersion and scouring. The detection of molten salt by the present invention is more in line with the working state of the molten salt, with high accuracy, and is intelligent and convenient.

The invention discloses a limb volume measuring instrument, which relates to the field of limb volume measurement and comprises a supporting groove, a sliding rod, a pulley trolley, a measuring frame, an infrared measuring grating and host equipment, the host equipment comprises a server; the server is used for sending a scanning signal to the infrared measurement grating; the infrared measurement grating is used for scanning the limb to be measured under the driving of the pulley trolley after receiving the scanning signal to obtain an infrared light signal; the server is also used for reading a group of infrared light signals scanned by the infrared measurement grating at set time intervals and storing each group of infrared light signals according to a time sequence; after the infrared measurement grating completely scans the limb to be measured, the length and width data of the section of the corresponding limb to be measured are calculated according to each group of stored infrared light signals; and inputting the length and width data of the to-be-measured limb section corresponding to all the infrared light signals into the volume error correction model to obtain corrected length and width data. The limb volume can be simply, conveniently and accurately evaluated.

The present invention relates to the technical field of laser spot welding machines, and proposes a laser spot welding machine for mold processing, comprising a workbench, a mounting bracket fixed to the rear side of the top wall of the workbench, a No. 1 electric slide rail fixed on the top wall of the mounting bracket, a vertical plate fixed to the top wall of the slider of the No. 1 electric slide rail, a No. 2 electric slide rail fixed to the side wall of the vertical plate, a crossbeam fixed to the side wall of the slider of the No. 2 electric slide rail, a laser spot welding machine body fixed to the front end of the crossbeam, a protective cover is provided on the outer side of the welding head of the laser spot welding machine body, and an air jet collection component, an eye protection component, a translation and lifting component and a purification component are arranged on the workbench; through the above technical scheme, the problem that most welding equipment in the prior art does not have integrated protective gas injection and exhaust gas treatment functions, which will threaten the health of operators and may also pollute the environment, and strong light, ultraviolet rays and infrared rays are generated during welding, and operators are prone to visual damage during the welding process is solved.

This invention proposes an online quality detection method for An Shen Bu Nao Liquid based on multispectral fusion. Addressing the bottlenecks of traditional detection techniques, such as long offline processing time, weak multi-index detection capabilities, and poor dynamic adaptability, this method achieves intelligent quality control throughout the entire production process. The device integrates ultraviolet-visible spectroscopy, near-infrared spectroscopy, and fluorescence spectroscopy modules. It constructs a multi-scale residual convolutional network to extract ultraviolet-visible spectral features, uses a two-way LSTM to analyze the fluorescence excitation-emission matrix, and employs dynamic baseline correction of the ultraviolet spectrum. This improves the detection accuracy of icariin content by 40% compared to traditional HPLC methods, and achieves a heavy metal residue detection limit of 0.05 ppm, providing an innovative solution for the intelligent production of traditional Chinese medicine preparations.

The invention discloses a citrus lossless identification method and system based on near infrared spectrum characteristic decoupling, and the method comprises the steps: obtaining near infrared spectrum data of a citrus sample, and coding the near infrared spectrum data to a submerged space to obtain spectrum characteristic representation; mapping the pooled features to a discrete citrus mode codebook by using a vector quantizer, and extracting mode features; performing multi-scale convolution operation on the spectral feature representation, and extracting fine-grained spectral features; performing orthogonal decomposition on the fine-grained spectral features into personalized features and invalid features through orthogonal decomposition; and according to the extracted pattern features and personalized features, citrus category identification is realized. The citrus lossless identification system integrates and realizes spectral feature representation, quantization, feature extraction and category identification functions. According to the method, the mode features and the personalized features of the citrus are extracted, the accuracy rate reaches 85.24% in 80 types of citrus identification tasks, and the method has a wide application prospect.

The invention provides an infrared femtosecond and infrared nanosecond laser combined machining system and a bionic machining method, and relates to the technical field of laser surface machining. According to the method, rapid rough machining and pre-structure construction are achieved through nanosecond laser, then fine modification and micro-nano structure regulation and control are conducted through femtosecond laser, the machining efficiency and the structure quality can be remarkably improved through the synergistic effect of the nanosecond laser and the femtosecond laser, meanwhile, the heat influence is reduced, high-precision and low-damage machining of the SiC surface is achieved, and the machining quality is improved. And a technical support is provided for constructing a surface structure with special functionality.

The invention relates to an infrared unfreezing method and device, an electronic device, a storage medium and a refrigerator, and the infrared unfreezing method comprises the following steps: obtaining the thickness of a to-be-unfrozen object, and determining a first infrared ray according to the thickness of the to-be-unfrozen object; obtaining the temperature of a partial freezing chamber in real time, and adjusting the wavelength of the first infrared ray according to the temperature of the partial freezing chamber to obtain a target unfreezing infrared ray; and heating and unfreezing the to-be-unfrozen object by adopting the target unfreezing infrared ray according to a preset heating control logic. Through the application, the problems of non-uniform thawing and low efficiency are solved, uniform thawing is realized, and the thawing efficiency is improved.

The invention discloses a deep learning wolfberry origin identification analysis model based on near infrared spectrum and Raman spectrum fusion. A wolfberry sample data set is acquired in a nondestructive manner through near infrared spectrum and Raman spectrum, spectral data is preprocessed, useful information is extracted, instrument noise is removed, and fluorescence background interference is reduced. The method comprises the following steps: firstly, constructing a cross-modal feature fusion model, namely, Multimodal Collastoration Attention Network (MCAN), and then, constructing a cross-modal feature fusion model, namely, the MCAN (Multimodal Collastoration Attention Network); according to the model, a spectrum branch network and a collaborative attention network are combined, information among different modes can be effectively integrated through feature level fusion, and the precision and robustness of Chinese wolfberry origin traceability identification are improved.

The invention discloses an infrared barrier PVB (polyvinyl butyral) intermediate film and a preparation method, and relates to the technical field of environmental functional glass materials, the intermediate film comprises 60-75 parts of polyvinyl butyral resin, 25-35 parts of a plasticizer, 0.5-3 parts of nano tin antimony oxide, 0.1-0.5 part of an ultraviolet absorbent, 0.05-0.2 part of an antioxidant and 0.1-0.5 part of a dispersant, the preparation method comprises the steps of nano infrared blocking agent pre-dispersion, mixing, twin-screw extrusion molding, tape casting cooling, surface treatment and the like, the surface-modified nano tin antimony oxide is introduced as the infrared blocking agent, infrared radiation at the wave band of 780-1300 nm is effectively blocked, the infrared blocking rate reaches 85% or above, the visible light transmittance is larger than 80%, the technology is mature and stable, the cost is low, and the preparation method is suitable for industrial production. The prepared intermediate film can be applied to building and automobile laminated glass, and the heat insulation and energy saving effects are achieved.