3221 results about "Test sample" patented technology

The invention relates to a deep learning model quantitative evaluation method and device and a storage medium, and the method comprises the steps: automatically determining a target quantification strategy through a preset mapping relation between a quantification strategy and model parameter information, does not need to manually debug parameters, carries out the model quantification of an optimized model based on a quantification tool chain configured by a configuration file, and improves the efficiency. According to the method, a quantitative model and a test sample are configured to obtain a deployment package, the deployment package comprises the quantitative model and the test sample, a reasoning index can be directly obtained on end-side equipment and compared with an original model index, objective quantitative evaluation is formed, and the method forms a standardized closed-loop process from model structure optimization, strategy generation and tool chain configuration to end-side deployment and evaluation. Manual intervention of connection of all links is reduced, dependence of an existing method on artificial experience is broken through, model quantitative evaluation can be automatically achieved, and the model quantitative requirement of a deep learning model is met.

The invention provides an intrinsic solid-state ion diffusion coefficient test method and device based on optical imaging, and relates to the technical field of solid-state ion test.The method comprises the steps that a test sample is pretreated; carrying out a photoelectric test on the test sample; placing the electrochemical cell in an optical imaging system, connecting the electrochemical cell with an electrical testing system, applying a constant voltage deviating from a steady state to the monodisperse electrode plate, collecting the real-time change condition of the material under the modulation of the constant voltage in real time through an industrial camera, and outputting relational data of optical intensity and time; analyzing the diffusion path through in-situ characterization equipment, and obtaining an ion diffusion path corresponding to the to-be-detected ion battery electrode material particles in combination with dynamic information provided by an optical imaging system; and determining the characteristic time according to the relational data of the optical intensity and the time, and determining the intrinsic solid ion diffusion coefficient corresponding to the test sample by combining the characteristic time and the ion diffusion path corresponding to the to-be-tested ion battery electrode material particles.

A true triaxial time-dependent dynamic disturbance test device for a deep-buried hard rock includes a rigid static load loading assembly and a rigid dynamic load loading assembly. The rigid static load loading assembly applies a static load to a rock test sample, and the rigid dynamic load loading assembly applies a disturbance stress to the rock test sample. Coupled application and control of a long time-dependent static force and a disturbance stress of the rock in a true triaxial disturbance stress state are achieved. The rigid dynamic load is provided with a disturbance rod in a loading direction, and a “point” disturbance mode is converted into a “face” disturbance mode through the disturbance rod, a disturbance hole and a test sample clamp, so that a disturbance mode more suitable for an actual working condition on site is provided while power consumption is reduced and time-dependent dynamic disturbance is achieved.

The invention discloses a motor current fault diagnosis method based on a de-noising diffusion probability model, and belongs to the technical field of mechanical equipment state monitoring and fault diagnosis, and the method comprises the steps: obtaining an original motor current signal sample, carrying out the wavelet transformation, obtaining a time-frequency grayscale image, and obtaining a time-frequency grayscale image; dividing the sample into a training sample used for diffusion model training and a test sample of a fault diagnosis model; constructing a diffusion model DDPM, carrying out training by adopting the training sample, and generating a pseudo sample based on the trained diffusion model; constructing a fault diagnosis model MAF-Cnet, and training the MAF-Cnet based on the training sample and the pseudo sample to obtain the trained MAF-Cnet; and inputting a test sample into the trained MAF-Cnet for diagnosis to obtain a motor current fault diagnosis result. The method solves the core problem that the generalization ability of the diagnosis model is insufficient due to scarcity of motor fault samples, improves the diagnosis accuracy, and is wider in application scene.

The invention discloses a multi-modal large language model training method and system, and relates to the technical field of multi-modal large model data processing, and the method comprises the following steps: training a first large model through querying a question set, a positive sample and a hard negative sample; inputting the plurality of test samples into the trained first large model, and generating a plurality of second answers based on a first preset prompt; sorting each test sample based on the plurality of second answers, and retrieving to obtain a previous candidate multi-modal document related to the test question; and training the second large model through the test problem and the corresponding previous multi-modal document. According to the method, the positive samples and the hard negative samples are jointly used for training, the model is forced to capture the fine-grained semantic boundary of correlation judgment through a contrast learning mechanism, the distinguishing capacity of the model for difficult samples is remarkably improved, and the mistaken arrangement phenomenon is avoided.

The invention discloses a multi-axis multi-field coupling fatigue test device, and belongs to the technical field of material testing. The device comprises a rack, a multi-axis loading unit, an environment simulation unit, a dynamic sealing interface and a cooperative control system. The core of the device is that the test sample can simultaneously bear the coupling effect of multi-axis mechanical stress (pulling, pressing and twisting) and various environmental fields (such as high and low temperatures, corrosive media, illumination and electric fields) through a highly integrated design; and the dynamic sealing problem of the actuator rod in the environment box is solved through an innovative dynamic sealing interface (such as labyrinth-magnetofluid composite sealing). According to the invention, accurate linkage control of multiple channels is realized through the cooperative control system, the service environment of key parts under actual complex working conditions can be truly simulated, and a reliable test platform is provided for failure analysis and life prediction of products under multi-field coupling.

The invention discloses a hyperspectral image open set classification method based on prototype similarity distribution modeling. The method comprises the following steps: acquiring a training sample set and a test sample set; initializing a category prototype on the unit hyper-sphere and forming uniformly distributed prototype spaces through iterative optimization; training a deep learning network based on a joint loss function, guiding the samples to form high-discrimination similarity distribution in a prototype space, and calculating similarity vectors of the samples and prototypes of each category; training the OCSVM by using the similarity vector of the known class sample to learn the distribution boundary of the known class in the prototype space; in the test stage, a test sample is input into the deep learning network, the similarity vector of the test sample is calculated, and whether the test sample belongs to a known class or not is judged by the OCSVM; and if the class is a known class, determining the fine-grained classification according to the class prototype of the maximum similarity component in the similarity vector. According to the method, the interpretability and generalization of the open set classification model are enhanced, and the misrecognition risk between similar categories is reduced.

The invention discloses a sealing structure for a test, and an aging test and leakage rate detection method and system. The sealing structure comprises a tested sealing assembly and an air cavity assembly integrated in the tested sealing assembly, and serialized test samples with different filling rates and compression rates are formed by regulating and controlling the groove width and the groove depth of a sealing ring mounting groove so as to simulate various actual sealing states. An aging test adopts a multi-factor coupling accelerated aging method, and a plurality of material characterization means are combined to systematically evaluate the evolution rule of the sealing performance. According to leakage rate detection, static pressure maintaining and dynamic pressure adjusting dual-mode measurement is achieved through an air tightness detector, and the sealing characteristics under different pressure conditions are comprehensively reflected. According to the invention, direct and systematic detection of the leakage rate of the sealing structure is realized, and an innovative solution is provided for sealing performance evaluation and service life prediction of electrical equipment.

The invention belongs to the technical field of traditional Chinese medicine detection, and discloses a detection method for a characteristic spectrum of Uighur medicine Kspecific heat extracting tablets and application of the detection method. The detection method provided by the invention comprises the following steps: preparing a test sample solution by taking a 75% methanol solution as a solvent, and determining the test sample solutions prepared from different batches of test samples by adopting ultra-high performance liquid chromatography to obtain a characteristic spectrum; the chromatographic conditions of the ultra-high performance liquid chromatography are as follows: a chromatographic column is a C18 chromatographic column, a mobile phase A is acetonitrile, a mobile phase B is a 0.1 vol% phosphoric acid solution, and the flow velocity is 0.25-0.35 mL / min; the detection wavelength is 0 to 18 minutes, and the wavelength is 327 nm; the time is 18 to 28.5 min, and the wavelength is 254 nm; when the time is 28.5 to 37.6 minutes, the wavelength is 242 nm; after 37.6 to 70 minutes, the wavelength is 210 nm; the sample size is 1 microliter, the column temperature is 20-40 DEG C, and gradient elution is carried out. The invention provides a more scientific identification and quality control method for the Kespecific heat tablets.

The invention relates to the technical field of high-pressure precise physical experiments, in particular to an irradiation sample table. A to-be-tested sample is placed on the sample table, and the second gland and the first gland are detachably connected to the sample table in sequence, so that the holes in the first gland and the second gland correspond to the to-be-tested sample. The size of the hole is smaller than the diameter of the to-be-detected sample, so that the radiation area is smaller than the diameter of the to-be-detected sample; moreover, the second gland is a ceramic gland and has good insulation performance, so that the situation that plasma directly bombards a sample table to generate an edge effect during irradiation, and adverse effects on judgment and separation of real particle flow density distribution borne by the surface of the to-be-detected sample are generated can be avoided. When samples of different sizes need to be replaced, only the first gland and the second gland which are provided with the open holes matched with the samples in size need to be replaced, operation is easy and fast, cost is low, and the experiment continuity and efficiency cannot be affected.

The invention discloses a nonlinear dielectric constant measuring system based on a scanning nonlinear dielectric microscope, comprising: a probe, one end of which forms a micro capacitance Cs with the upper surface of a sample to be tested; the low-frequency signal module applies a low-frequency signal to the lower surface of the to-be-tested sample; the other end of the probe is connected with an LC resonance circuit in the probe module; the output end of a signal amplifier in the probe module is connected with a frequency modulation demodulator; the frequency modulation demodulator demodulates the frequency signal modulated by the probe module, removes a high-frequency carrier signal, inputs the demodulated signal into the lock-in amplifier, and detects, separates and amplifies a high-signal-to-noise-ratio signal by taking the frequency of a low-frequency signal applied to the sample to be tested as a reference frequency; according to the measurement system, the detection speed, the service life of the probe and the multi-material applicability can be remarkably improved, meanwhile, the thermal disturbance effect is inhibited, and a reliable solution is provided for accurate characterization of nano electronic devices.

The invention provides a deep learning-based critical-size small-angle X-ray scattering data processing method, which comprises the following steps of: generating simulation data based on a physical model, constructing an omega-qxz two-dimensional image, and pre-training and storing a neural network model by utilizing the simulation image; small-angle X-ray scattering experiment data of the sample is collected, an experiment omega-qxz two-dimensional image is constructed, and a neural network model is trained and fine-tuned by using a training method during testing; and outputting critical dimension parameters according to the experimental image by using the trained and fine-tuned neural network. According to the method, the neural network is used for processing the two-dimensional image data, the key size parameters of the sample are rapidly predicted from the scattering data obtained from the sample at a small number of rotation angles, scattering signals at different angles do not need to be reconstructed to form a two-dimensional image under a qx-qz coordinate system, the testing speed is increased, and the testing efficiency is improved. And the time consumption of sample testing and data processing is reduced.

The invention belongs to the technical field of asphalt mixed material experiments, and particularly relates to test equipment for measuring the high-temperature stability performance of an asphalt mixed material, the test equipment comprises a test platform, an environment control module and a data acquisition module, the test platform comprises a test bracket, samples to be tested are uniformly arranged on an inclined adjusting plate on the test bracket, and the environment control module is connected with the data acquisition module; the data acquisition module comprises a laser displacement sensor, and the laser displacement sensor is arranged at the position, corresponding to the to-be-tested sample, of the adjusting plate; according to the invention, the laser displacement sensor located in the protection box is started at regular time, and the relative displacement change between the laser displacement sensor and the to-be-tested sample is calculated according to the change of the reflected laser, so that the actual deformation displacement of the to-be-tested sample is accurately measured and recorded and is transmitted to external computer equipment; and real-time and accurate data acquisition of the deformation degree of the to-be-tested sample is realized.

The invention provides an in-vitro skin blood vessel immune model as well as a preparation method and application thereof. THP-1 human monocyte leukemia cells are induced to be differentiated into M0 type macrophages, and then the M0 type macrophages are respectively induced into M1 type macrophages and / or M2 type macrophages; then co-culturing the M1 type and / or M2 type macrophages and vascular endothelial cells to form a vascular immune model; finally, the 3D skin model is placed on the blood vessel immune model, external stimulation is conducted, and the in-vitro skin blood vessel immune model is constructed. The in-vitro skin blood vessel immune model can be used for repairing skin barriers, inflammation pathways, blood vessel metabolism, the expression level of genes or proteins related to extracellular matrixes, vascular endothelial cells, the proliferation and migration ability of activated macrophages and the like. The seven feature dimensions of the physiological structure feature of the skin model are used for screening the to-be-tested sample and exploring the action mechanism, and the method has the characteristics of rapidness, multiple screening dimensions, high accuracy, low construction difficulty, low cost and high universality.

The invention relates to a defect detection device and method, computer equipment and a computer readable storage medium. The device comprises an imaging subsystem, a structured light illumination subsystem and a data fusion processing unit. The structured light illumination subsystem comprises a coupling device, and the coupling device is used for coupling an illumination light path of the structured light illumination subsystem into an imaging light path of the imaging subsystem; the imaging subsystem is used for synchronously acquiring a light field image of the tested sample under the condition that the structured light illumination subsystem projects the structured light to the tested sample, and the light field image comprises light field depth data and a structured light image; and the data fusion processing unit is used for fusing the light field depth data and the structured light image to reconstruct the three-dimensional morphology of the detected sample, and generating a defect detection result of the detected sample based on the three-dimensional morphology. By adopting the device, the reconstruction precision of the three-dimensional morphology can be improved, and tiny pit defects on the surface of a transparent / weak texture sample can be identified.

The invention relates to the technical field of medicine analysis and detection, and particularly discloses a method for detecting the content of protein in a 3D-MLA adjuvant, which comprises the following steps: (1) preparing a test solution: firstly dissolving the 3D-MLA adjuvant by using a hydrochloric acid solution, performing 300-600W ultrasonic treatment for 2-5min, centrifuging 10000-13000g for 4-7min, and taking supernate; adding a compound enzyme into the supernate for enzymolysis; then adding concentrated hydrochloric acid, uniformly mixing, hydrolyzing for 20 + / -2 hours in a nitrogen atmosphere at 105-115 DEG C, blow-drying, and dissolving with a hydrochloric acid solution to obtain a test sample stock solution; adding a derivatization reagent into the test sample stock solution, and carrying out derivatization treatment, extraction and dilution to obtain a test sample solution; and (2) high performance liquid chromatography determination. According to the detection method provided by the invention, the residual quantity of the host cell protein in the 3D-MLA can be rapidly and accurately determined, and the quality control of the 3D-MLA is realized.

The invention discloses a helium detection device and method for the airtightness of a bonding interface and a bonding microcavity, and belongs to the field of the airtightness of MEMS devices. The device comprises a sample holder, a mass spectrometer, a molecular pump, a mechanical pump and other components, helium is used as detection gas, and the helium is leaked only through a bonding interface through a customized and processed test sample and sample holder structure, so that the interference of other leaks in a non-bonding area on a test result is avoided. The testing method comprises the steps of sample introduction, vacuum pumping, helium filling, signal detection, dynamic and steady state testing and the like, the helium leak rate is monitored in real time by combining with a helium mass spectrometer, and the minimum leak detection rate is that high-sensitivity air tightness detection is ensured. According to the method, the airtightness of the bonding interface can be independently evaluated, and a reliable basis is provided for development of a novel bonding process. The method is suitable for air tightness detection and evaluation of the bonding process in MEMS device processing and manufacturing.

The invention provides a hyperspectral image reconstruction method based on space-spectral characteristic fusion. The method comprises the following implementation steps: acquiring a training sample set and a test sample set; constructing a reconstruction network model based on space-spectral characteristic fusion; performing iterative training on the reconstructed network model; and obtaining a reconstruction result of the hyperspectral image. According to the invention, the spatial-spectral characteristic guiding network in the encoder carries out spatial-spectral characteristic extraction on input characteristics, the spatial-spectral characteristic guiding network in the bottleneck layer carries out spatial-spectral characteristic fusion, the spatial-spectral characteristic guiding network in the decoder carries out spatial-spectral characteristic detail recovery, spatial structure information and spectral distribution information in the input characteristics are fully extracted, and the spatial-spectral characteristic guiding network in the decoder carries out spatial-spectral characteristic fusion. The spatial features and the spectral features cooperatively participate in the reconstruction process, and cooperative recovery of the spatial information and the spectral information of the hyperspectral image is realized, so that the reconstruction precision of the hyperspectral image is effectively improved.

The invention discloses a wheat yield remote sensing prediction method combining phenological parameters, which comprises the following steps: S1, performing field observation in a key growth period of winter wheat, synchronously collecting canopy hyperspectral reflectivity data, leaf area index (LAI) and SPAD value of each observation sample point, and recording wheat grain yield of the corresponding sample point; s2, performing noise reduction preprocessing on the acquired canopy hyperspectral data, and extracting sensitive spectral parameters by combining principal component analysis (PCA) and correlation analysis methods; s3, taking the sensitive spectrum parameters, LAI and SPAD values as independent variables, taking the wheat grain yield as a dependent variable, and adopting partial least squares regression (PLSR) to construct a multivariable yield prediction model; and S4, performing wheat yield prediction on an independent test sample or regional scale remote sensing data by using the trained model. The method overcomes the defect that only yield sensitive spectrum parameters are used for predicting the effect, and accurate estimation of the model is achieved.

The invention discloses a large language model security testing method based on evolutionary dynamic adversarial attacks, and relates to the technical field of security testing. Comprising the following steps: 1, creating a dynamic attack generation engine, constructing an adversarial evolution architecture by utilizing the dynamic attack generation engine, and generating a test sample based on the adversarial evolution architecture for a large language model security test; the method comprises the following steps: 1, establishing a large-scale language model, 2, calculating investigation parameters of harmlessness, honesty and helpfulness based on a Constancy AI principle, calculating a security alignment gap index SAGI by using the investigation parameters, and intelligently judging a value drift condition of the large-scale language model according to the security alignment gap index SAGI, 3, establishing a multi-modal joint defense engine, and carrying out intelligent judgment on the value drift condition of the large-scale language model according to the value drift condition of the large-scale language model. Steganalysis, syntax tree analysis and audio anomaly detection functions are integrated, and all-around threat detection coverage of texts, codes, images and voices is carried out on a detected large language model.

The invention discloses high-voltage wiring harness connector terminal fretting wear test equipment which comprises an equipment base, a high-frequency vibration exciter and a lifting platform are mounted on the equipment base, and a mounting platform is arranged at the top of the lifting platform; a left test piece mounting loading mechanism and a right test piece mounting force measuring device are arranged on the mounting platform. The device can realize a multi-physics field coupling test, current-carrying friction and arc erosion of an electric contact surface in a real working condition are simulated by combining a current load with high-frequency mechanical wear, and the electric-mechanical coupling wear characteristic of a new energy automobile charging terminal is accurately reproduced. The device supports the dynamic adjustment of normal load, simulates the contact pressure change of the left and right test samples of the test terminal in the plugging process through the test sample pin and the test sample jack, can be used for researching the relevance between the fretting wear and the contact resistance of the test terminal under different loads, and realizes the test of the terminal under complex working conditions in a laboratory environment. The test process is convenient, efficient and accurate, and the test cost and the test period are greatly reduced.

The invention provides a multi-mode coupling flood forecasting method and system based on a gated cross attention mechanism, and relates to the technical field of flood forecasting. The method comprises the following steps: acquiring high-precision rainfall monitoring information, numerical forecast rainfall information and hydrological station session flood excerpt information of a predetermined drainage basin; constructing training, verification and test samples; a spatial-temporal feature coding module based on a convolutional neural network and Transform is constructed, and rainfall information is fully mined; a time sequence feature coding module based on a time sequence convolutional network is constructed, and runoff information is fully mined; constructing a multi-modal feature fusion module based on a gated cross attention mechanism, and fusing different modal features mined by the time sequence and spatial-temporal feature coding modules; and constructing a flood prediction module based on a self-attention mechanism to perform flood prediction. According to the invention, the accuracy and reliability of flood forecasting are improved based on the multi-modal feature fusion technology.

The invention discloses a Transform low-light image enhancement method based on dynamic four-priori, and the method comprises the steps: obtaining a low-light image data set, and dividing the low-light image data set into a training sample set, a verification sample set and a test sample set; a Transform low-light image enhancement model DQPGT based on the dynamic four priori is constructed; inputting the training sample set into a DQPGT model to train the model, and verifying the model through the verification sample set to complete training of the DQPGT model; inputting a test sample set into the trained model for image enhancement and then outputting the test sample set; according to the method, spatial and frequency domain feature representation is collaboratively optimized on the basis of dynamic four-priori estimation and a frequency domain channel attention mechanism, so that the problems of uneven exposure, noise amplification and color distortion of an existing method based on the Retinex theory under a complex real illumination condition are solved.

The invention belongs to the field of geotechnical engineering, and particularly relates to a sand liquefaction discrimination method based on CPTU and weighted nonlinear similarity. Comprising the following steps: collecting sand CPTU data containing liquefaction and non-liquefaction cases, and calculating a cyclic stress ratio (CSR) as a derivative feature; preprocessing and standardizing the data and dividing the data into a training set and a test set; respectively calculating prototype vectors of liquefaction and non-liquefaction categories based on the training set; by analyzing the influence of disturbance of each characteristic value on the similarity of the sample and the prototype vector, calculating a characteristic weight to quantify the importance of the characteristic weight; constructing a weighted radial basis kernel function fused with feature weights, calculating the nonlinear similarity between the to-be-tested sample and the two types of prototype vectors, and converting the nonlinear similarity into a liquefaction probability; and based on the confidence coefficient of the information entropy quantification prediction result, uncertainty evaluation is realized. The method has high discrimination precision, strong interpretability and reliable uncertainty quantification capability, and effectively overcomes the defects of strong subjectivity and black box of a machine learning model in a traditional experience method.

The invention relates to a hyperspectral image classification method based on BoostFormer double-view perceptual feature fusion, and belongs to the technical field of information processing. The method comprises the following steps: S1, hyperspectral image acquisition and data preprocessing; s2, inputting the training sample into a BoostFormer dual-view perceptual feature fusion network for learning; s3, calculating loss and updating model parameters; s4, after training is completed, the test samples are classified, and a test result is obtained. The performance of the method provided by the invention on a hyperspectral image classification task is superior to that of other hyperspectral image classification algorithms.

The invention discloses a self-supervised defect representation learning method in an industrial scene, and the method comprises the steps: constructing a normal sample data set of an industrial product, carrying out the preprocessing, and obtaining a preprocessed normal sample; inputting the preprocessed normal samples into a multi-scale feature pyramid double-branch encoder network, extracting and fusing global semantic features and local detail features, and obtaining fused features; carrying out self-supervised training through a progressive temperature regulation contrast learning mechanism by utilizing the fusion features so as to learn feature distribution of normal samples; the fusion features are decoded and reconstructed through a semantic consistency reconstruction network, and reconstruction loss is calculated to optimize feature representation; and calculating a comprehensive abnormal score of a to-be-detected sample based on the trained model, establishing a dynamic threshold according to statistical distribution of the comprehensive abnormal score, and judging defects by comparing the comprehensive abnormal score of the to-be-detected sample with the dynamic threshold. According to the invention, the generalization capability and the detection precision of the industrial quality inspection system are obviously improved.

The invention provides an intrinsic solid-state ion mobility testing method and device based on optical imaging, and relates to the technical field of performance testing of solid-state ion electrolyte materials.The method comprises the steps that a test sample pool is assembled; performing a power-on test on the test sample pool; extracting size information of solid electrolyte powder particles to be detected from the optical image; extracting a time-varying curve of the optical intensity of the solid electrolyte powder particles to be detected from the optical image, and determining time information of the optical response of the particles from beginning to balance in the voltage variation process based on the curve; determining the electric field intensity under the test condition according to the magnitude of the applied voltage and the total thickness of the thin-layer insulating substrate; and determining the intrinsic ion mobility of the solid electrolyte powder particles to be detected according to the size information, the time information and the electric field intensity.

The invention discloses a model security test method, device and equipment and a storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining a test sample composed of a test image and a corresponding target detection label, and a to-be-tested visual target detection model used for executing a target detection task, and generating an adversarial sample based on the test sample by using the to-be-tested visual target detection model, respectively inputting the test sample and the adversarial sample into the to-be-tested visual target detection model to obtain a corresponding prediction result so as to determine a target test score of a target neuron in the to-be-tested visual target detection model, and generating a security test result. By comparing and quantizing the neuron fraction through the results of the adversarial sample and the test sample, the sensitive neuron in the model is determined, the model safety can be determined more accurately, and the anti-interference capability, safety and stability of the model are improved.

A blood cell analyzer (100) and a blood cell analysis method (300). The blood cell analysis method (300) comprises: sucking a blood sample to be tested; preparing a first determination sample and obtaining first optical information; preparing a second determination sample and obtaining second optical information; generating a first leukocyte classification scatter diagram and a second leukocyte classification scatter diagram; determining a first counting result of neutrophils of the to-be-detected blood sample based on the first leukocyte classification scatter diagram and the second leukocyte classification scatter diagram; determining an actual distribution area of neutrophils from the first leukocyte classification scatter diagram; based on a reference distribution region of neutrophils of a reference leukocyte classification scatter diagram of the normal blood sample, determining a leaf-division nucleophore distribution region from the actual distribution region; determining a second counting result of the cells falling into the leaf-separating nucleogranulocyte distribution area; and based on the first counting result and the second counting result, determining a counting result of the rod-shaped nuclear granulocytes in the to-be-detected blood sample.

The invention discloses a bearing steel open-set classification system based on differential pulsed eddy current and a double-branch network, and the system comprises a differential collection module which is used for obtaining steel induction eddy current signal data and providing required original data; the data preprocessing module is used for carrying out signal preprocessing on the collected original data and improving the signal-to-noise ratio of signals; the model structure optimization module is used for fusing manually selected statistical features and machine learning features to form a double-branch network model structure, so that the signal classification precision and the discrimination capability of unknown bearing steel samples are improved; by calculating the confidence coefficient of a to-be-tested sample and the distance between the class center of the to-be-tested sample and each class center recorded by the model, which class is in known classes can be judged, and by setting a class center distance threshold, an unknown class can be effectively identified.