52 results about "Differential analysis" patented technology

This invention discloses a three-dimensional quantification method for transmission line defects based on RGB-D point clouds and PointNet++, aiming to solve the technical problem that traditional two-dimensional visual inspection can only qualitatively identify defects and cannot achieve accurate quantification of physical parameters. The method includes constructing a dual-modal data acquisition framework of RGB texture and dense point cloud, achieving pixel-level spatiotemporal alignment through an online calibration algorithm for heterogeneous sensors; building a semantic segmentation model based on PointNet++ to accurately segment key components of the transmission line; and establishing a differential analysis mechanism between the standard equipment 3D model and the measured point cloud model to automatically calculate key parameters such as icing thickness, hardware corrosion volume, and insulator damage area. This invention breaks through the limitations of traditional two-dimensional visual qualitative identification, realizing a leap from visual inspection to physical quantification of defects, filling the gap in three-dimensional quantitative assessment in the industry, and providing accurate data support for the digital operation and maintenance of transmission lines.

This invention proposes an early warning system and method for the impact of ecological submerged breakwaters on nearshore seabed morphology under tsunami wave action. The system includes: a feature extraction module that collects water pressure data inside and outside the breakwater and establishes a high-energy potential difference time window; a potential energy analysis module that retrieves data within the high-energy potential difference time window and performs differential analysis of the vertical gradient to generate the permeability per unit volume of the seabed surface; an erosion analysis module that obtains the loss coefficient and corrects the static critical initiation shear stress to establish a time-evolving erosion resistance threshold sequence; a concealment identification module that identifies concealed scour zones; and a graded early warning module that quantifies and accumulates the surplus characteristics of shear strength in concealed scour zones using sediment transport rates to obtain morphological loss characteristics, performs stability assessment of the breakwater support structure, and triggers graded collapse early warning. This invention achieves the identification and dynamic early warning of concealed scour risks under low flow velocity backgrounds by quantifying soil weakening caused by seepage.

The present application relates to the technical field of cloud computing and virtualization, and particularly to a cloud service resource scheduling system based on virtual machines; comprising benchmark construction, parallel differential analysis, feature matching and scheduling execution modules; the system collects real-time indexes to generate ideal values without interference, and calculates real deviation and theoretical deviation based on interference models; the core is to calculate the similarity of the two types of deviation in the feature space, and accordingly to distinguish resource contention and business increment, to perform cross-machine live migration when the similarity meets the standard, and otherwise to perform in-place expansion; the present application converts abnormal diagnosis into a pattern matching problem, accurately identifies the root cause of performance degradation, and effectively avoids misjudgment and invalid operation of traditional schemes.

The invention discloses a method for detecting potential electric leakage defects of an LED device, and relates to the technical field of semiconductor photoelectric device detection. Aiming at the technical pain point that recessive microdefects (such as a metal migration channel caused by a V-shaped pit and dislocation) cannot be identified by the existing single spectroscopic detection, the invention provides a detection logic of reference parameter acquisition, defect stress excitation, parameter acquisition after excitation and differential analysis and judgment. Specific stress is applied to primarily screened qualified products, ion migration at potential defects is excited in an oriented mode, and hidden electric leakage is dominated. Potential defective products are accurately rejected by calculating parameter drift distances (VFL, IR) before and after stress. The method is not only suitable for interactive screens, but also suitable for high-reliability scenes such as vehicle-mounted display and high-end illumination, and can effectively intercept hidden lamp beads causing'caterpillar 'failure or early light attenuation, thereby remarkably improving the product yield and the terminal reliability.

This invention discloses a comprehensive evaluation method for drought resistance of maize varieties, belonging to the field of maize variety evaluation technology. The method includes the following steps: acquiring multi-temporal, multi-band reflectance spectral data under continuous drought conditions, and combining this with the structural characteristics of the waxy coating on maize leaf epidermis to perform structural coupling modeling of the spectral data, constructing a time-series curve characterizing the dynamic changes in leaf wax distribution as a dynamic spectral monitoring baseline; performing differential analysis between consecutive time points based on the dynamic spectral monitoring baseline to identify spectral abrupt change points caused by leaf wax migration, and combining this with spatial consistency testing. This invention accurately identifies wax migration interference by constructing a dynamic spectral monitoring baseline and energy shift fingerprint, overcoming the assumption of spectral stability, and achieving high accuracy, adaptability, and consistency in evapotranspiration rate inversion under drought conditions through multi-level correction and closed-loop update strategies, thereby improving the scientific rigor and comparability of drought resistance evaluation.

This invention discloses a method for monitoring the production environment of nonwoven products, specifically relating to the field of production environment monitoring technology. The method involves collecting temperature, humidity, electrostatic intensity, and suspended particle concentration data from multiple areas of the production workshop to construct an environmental state sequence matrix; performing anomaly detection based on an environmental anomaly threshold model to extract a set of regional feature points; calculating dynamic risk factors by combining historical anomaly frequency and spatial proximity to generate a set of high-priority monitoring areas; constructing a regional environmental evolution model to predict parameter trends and identify potential secondary out-of-control areas; merging high-risk and secondary out-of-control areas to form a set of microenvironmental anomaly zones and implementing targeted environmental control; collecting a new round of data after control and performing differential analysis to evaluate the control effect. This method enables continuous identification and dynamic intervention of the workshop microenvironment, improving the production quality and environmental stability of nonwoven products.

This invention discloses a method, equipment, and medium for differential protection of phase-to-phase faults on 10kV overhead lines, relating to the field of power protection technology. The method includes: quantifying the instantaneous waveforms of three-phase current and voltage to obtain discrete sequences; obtaining the line current and line voltage change waveforms through point-by-point differential analysis; detecting the power direction angle offset using cross-phase detection; determining the fault direction by combining the initial polarity of the current jump; latching the trigger differential judgment start signal for a positive-direction fault; nonlinearly mapping the fault phase current with the transient characteristic duration to obtain a fault level coefficient; performing interval matching of the fault level coefficient based on a preset action delay level library; starting a delay timer according to the matched delay level; and performing a tripping operation after the delay timer expires to disconnect the faulty section of the overhead line. This invention improves the efficiency of fault isolation and achieves accurate graded protection and rapid section clearing for phase-to-phase faults on 10kV overhead lines.

This invention belongs to the field of biomedical technology, specifically relating to the application of plasma exosome biomarkers in the diagnosis and early warning of diabetic bladder dysfunction. This invention pioneered the use of 4D-DIA proteomics technology to conduct a comprehensive differential analysis of plasma exosome proteins from patients with diabetic bladder dysfunction (DBD). It successfully identified protein components with significantly different expression levels in the plasma exosomes of DBD patients, demonstrating significant value as potential diagnostic biomarkers. This invention is the first to propose that the detection of a four-protein combination in plasma exosomes (FCGR3B+THBS1+HLA-F+ATP5PB) can serve as a biomarker for diabetic bladder dysfunction (DBD) and also as an early warning indicator for the disease. This protein combination achieved an overall accuracy of 83.3% in identifying DBD.

The present disclosure provides a model compilation security evaluation method and system based on structural detection and differential analysis, relating to the technical field of compilation security evaluation, comprising: calling a compiler for compilation to generate intermediate representations and compilation products in each stage; performing single-file static detection on the intermediate representation graph file to identify neural network backdoor patterns embedded in the form of mathematical calculation structures in the graph; reading the intermediate representation graph file before and after operator fusion, performing differential analysis on the intermediate representation graph file before and after fusion, and identifying abnormal operators or data streams injected in the fusion stage; performing differential analysis on the kernel source code file and the machine code disassembly file to identify malicious tampering injected in the code generation or binary compilation stage; and synthesizing the detection results of multiple identification processes to perform final security evaluation on the model compilation process and generate a structured evaluation report. The present disclosure improves the identification capability of abnormal structures or potential backdoor logic hidden in the model compilation stage.

The embodiment of the present disclosure provides a gene variable splicing isomer batch effect deduction method, which belongs to the technical field of data identification and specifically comprises the following steps: acquiring a GTEx transcript count matrix as a parameter estimation reference data source; estimating a negative binomial distribution parameter and sampling to generate a new count matrix; injecting a batch effect and a differential transcript to use a DTU signal to construct a simulation data set; performing an ILR space countermeasure batch process on the simulation data set, that is, constructing an intragenic composition vector with a gene as a unit and performing ILR transformation, learning a batch discrimination direction in the ILR space with a batch variable as a supervision signal and projecting to remove, and then recovering to a count matrix after batch removal through inverse ILR transformation; and running a differential analysis method before and after batch removal to obtain a result table and comparing and evaluating. Through the scheme of the present disclosure, false positives are effectively reduced, the accuracy and robustness of DTU detection are improved, and the scheme is suitable for transcriptome data analysis.

This invention discloses a method and device for monitoring patchy fog based on a visible light image sensor and a point laser, belonging to the field of atmospheric visibility monitoring technology. The device includes a linearly calibrated visible light image sensor, a point laser emitter mounted parallel to the sensor's optical axis, and a synchronization control and data processing unit. The core of the method lies in: synchronously controlling and alternately acquiring image pairs of the laser being on and off, performing differential analysis to eliminate background interference, and obtaining a pure laser atmospheric scattering signal; integrating along the laser spot normal direction to obtain an initial brightness profile curve; based on the fixed spatial pose relationship between the sensor and the laser, mapping the brightness curve in image coordinates to a corrected brightness curve in real spatial distance coordinates through a geometric correction model; calculating the mean and variance of the corrected brightness curve, and calibrating it using standard instruments; and finally simultaneously obtaining the average path visibility value and the patchy fog intensity index. This invention is low-cost, has strong anti-interference capabilities, and can effectively achieve quantitative identification and spatial positioning of patchy fog, making it suitable for large-scale deployment in scenarios such as highways.

The application relates to the field of intelligent electric meter detection, in particular to an intelligent electric meter power consumption anomaly detection method and system based on mode classification, which comprises the following steps: obtaining historical current sequences and to-be-detected current sequences of each user, performing multi-scale equal-length adaptive segmentation and frequency band division on the historical current sequences, combining user intra and inter power consumption mode differences to calculate power consumption tolerance of each frequency band, performing EMD decomposition and feature extraction on the historical current sequences, calculating mode classification distances based on power consumption tolerance and clustering to obtain power consumption mode clusters, and finally comparing distances between the to-be-detected sequences and the mode clusters, calculating anomaly scores and determining power consumption anomaly states. Through multi-scale equal-length adaptive segmentation, frequency band differential analysis on the historical current sequences, and combination of user intra and inter power consumption mode differences, tolerance and anomaly scores are constructed, precise and personalized anomaly detection of user power consumption is realized, and detection accuracy and efficiency are improved.

This invention relates to the field of non-destructive testing technology for power equipment, and discloses a transformer coil fault detection system based on ultrasonic reflected echoes. It primarily addresses the problem of severe multipath interference in ultrasonic signals caused by the multi-layered structure inside the transformer, making it difficult to accurately extract fault features of specific coils. This system transmits and receives ultrasonic waves through components on the outer wall of the tank; it calculates the theoretical time of flight using a three-dimensional acoustic mapping model of the transformer and constructs a target time window; it performs time-domain truncation on the original echo, directionally stripping away interference to extract the effective reflected echo from the specific target coil; it extracts the acoustic feature parameters and obtains the feature difference vector by differential analysis with baseline features; finally, it determines the type of physical deformation of the coil and outputs the quantified deformation degree, achieving non-invasive, high-precision fault diagnosis.

This invention discloses a scalable distributed multi-chamber open-path photosynthetic measurement system and method, relating to the field of plant physiological and ecological measurement technology. The system includes a main unit and at least two distributed chambers. The main unit centrally manages gas supply, concentration mixing, flow control, and high-precision concentration differential analysis. Each chamber is an independent unit, equipped with an adjustable LED light source, thermoelectric temperature control module, multiple types of sensors, independent power supply, and wireless communication module. An open path is formed through the gas supply pipeline and the return pipeline. The core of the method lies in the main unit polling and sampling the return gas from each chamber according to a preset program through a multi-channel automatic switching valve group. Combined with a "flushing stabilization" strategy, a single gas analysis unit is used to achieve high-precision, high-throughput parallel measurement of multiple chambers. This invention solves the problems of low throughput and high cost of traditional open-path instruments, achieving a balance between high precision and high throughput, and supports in-situ deployment in the field and remote monitoring.

This invention discloses a method and system for predicting faults in wind power equipment, belonging to the field of wind power generation technology. The method includes: real-time acquisition of operating parameters of the wind turbine and measured response signals of the blade structure; construction of an aerodynamic load-structural response mapping model, and real-time calculation of the theoretical reference response signal of the blade based on the operating parameters; dynamic phase alignment and point-by-point differential operation on the measured response signal and the theoretical reference response signal to generate a residual signal after removing environmental load components; extraction of time-domain and frequency-domain multi-dimensional sensitive features from the residual signal, and fault early warning through a two-level discrimination mechanism of sliding window trend analysis and sudden threshold detection. This invention, by combining a physical model with signal differential analysis, effectively suppresses strong random environmental interference such as wind speed fluctuations and yaw loads, significantly reducing false alarms and missed alarms, and improving the sensitivity and reliability of early warning for identifying weak structural damage such as early cracks, icing, and loose bolts.

The application relates to the technical field of image recognition, and discloses a metal plate defect point detection method and system. The method comprises the following steps: obtaining a surface image of a target plate as an original image of the target plate; performing noise suppression on the original image to obtain an enhanced image of the target plate; performing texture direction feature extraction on the enhanced image to obtain a texture trend distribution map of the target plate; performing differential analysis on the texture trend distribution map and a preset standard texture direction map to obtain direction abnormal points of the enhanced image; performing gray scale statistics on the direction abnormal points to obtain a local gray mean value of the target plate, and determining a pixel point with a gray value lower than the local gray mean value as a suspected defect point; and performing morphological clustering on the suspected defect point to obtain a defect area of the target plate. The application can improve the defect point detection efficiency of the metal plate.

The application relates to the technical field of metabolomics analysis, and discloses an AI-assisted metabolomics difference analysis method, which comprises the following steps: performing data preprocessing on biological groups to obtain preprocessed data; identifying user input instructions, obtaining corresponding biological groups and output content categories based on the user input instructions; performing difference analysis on the preprocessed data corresponding to the biological groups through an AI analysis model to obtain difference analysis results and output content categories; and generating a complete analysis report according to the difference analysis results and the output content categories. Through the data preprocessing process, the application can separate and remove system variation (interfered by non-biological factors such as batch effect, instrument drift and sample collection time difference) from the data, highlight the real group difference, avoid false positive or false negative results, and ensure the accuracy of the analysis results when the AI performs difference analysis.

This invention provides an intelligent monitoring method and device for cold chain transportation packaging, relating to the fields of cold chain logistics and intelligent sensing technology. The method includes the following steps: acquiring infrared temperature distribution images of the surface of the transported goods, performing spatial discretization and weighted calculations to obtain a temperature set of surface assessment points; based on the surface temperature set, obtaining the temperature change sequence of internal monitoring points through inversion calculation using a heat conduction model; performing temporal differential analysis on the temperature change sequence to extract a dynamic temperature rise rate index characterizing overheating risk; integrating the dynamic temperature rise rate indices of all monitoring points, and generating a global risk index characterizing the overall heat load situation through spatial weighted fusion; triggering different levels of early warning information based on the comparison results of the global risk index and preset classification thresholds. This invention achieves non-invasive, dynamic, and forward-looking monitoring of the internal temperature of transported goods, significantly improving the accuracy and timeliness of risk warnings in the cold chain transportation process.

This invention provides an infrared detection system and method for thermal leakage on complex metal surfaces, comprising: a high-temperature steam generating unit; and an infrared imaging unit including an infrared camera capable of capturing infrared image sequences of a test sample. The high-temperature steam generating unit sequentially injects high-temperature steam into the test sample to heat it, with different pressures in the sample during the two steam injections. During each steam injection, the infrared camera captures a sequence of infrared images of the test sample. The infrared detection system identifies leakage points in the test sample by performing differential analysis on the infrared image sequences from the two heating processes. This invention effectively eliminates interference from different emissivities on the surface of the test piece, achieving accurate and convenient detection of thermal leakage on complex metal surfaces.

This invention belongs to the field of building damage detection, specifically a dynamic building damage detection method based on multimodal image processing. This invention introduces infrared thermal imaging as a second modality to achieve physical differentiation and fully utilizes the complementary characteristics of the two modalities. It also introduces continuous frame differential analysis and damage motion vector fields, elevating damage detection from static snapshots to dynamic process observation, providing direct evidence for trend prediction and safety early warning. Furthermore, this invention leverages the physical independence of infrared thermal imaging to enhance robustness. Finally, it introduces thermodynamic features to achieve indirect perception of depth information, enabling indirect perception and quantitative evaluation of damage depth, making safety assessment results more comprehensive and reliable. Simultaneously, by constructing a unified spatiotemporal joint modeling framework, it effectively eliminates the disconnect between detection and tracking, achieving end-to-end collaborative optimization.

This application relates to the field of data processing technology, specifically providing a method, apparatus, storage medium, and electronic device for incremental data synchronization. The method may include: performing differential analysis on business resource information to be processed and a source data knowledge graph to obtain the data type of each type of resource object in the business resource information to be processed; wherein the data type includes newly added data, updated data, or deleted data; constructing a data set matching the data type of each type of resource object; wherein the data set includes a newly added set, an updated set, or a deleted set; and updating the source data knowledge graph using the data set to obtain an updated knowledge graph. Embodiments of this application can improve the efficiency of incremental data synchronization.

A method and apparatus for identifying a difficult-to-differentiate bacterial species, a device, and a medium, relating to the technical field of microbial identification, and comprising: acquiring a mass spectrometry identification result sent by a source mass spectrometer, determining whether a difficult-to-differentiate bacterial species is present in the mass spectrometry identification result, and if the difficult-to-differentiate bacterial species is present in the mass spectrometry identification result, determining whether there is a fine-grained differential analysis model corresponding to a type of the difficult-to-differentiate bacterial species in a model library; if there is a fine-grained differential analysis model corresponding to the type of the difficult-to-differentiate bacterial species in the model library, invoking the fine-grained differential analysis model to analyze and identify the difficult-to-differentiate bacterial species so as to obtain a model analysis and identification result; if there is no fine-grained differential analysis model corresponding to the type of the difficult-to-differentiate bacterial species in the model library, using a dynamic matching algorithm to perform dynamic computational identification on the difficult-to-differentiate bacterial species, or using the dynamic matching algorithm and invoking the fine-grained differential analysis model to perform dynamic computational identification on the difficult-to-differentiate bacterial species to obtain a dynamic computational identification result, thereby improving matrix-assisted laser desorption time-of-flight mass spectrometer accuracy and efficiency with respect to identifying similar microorganisms.

The application discloses a method for evaluating the curative effect of traditional Chinese medicine in colorectal cancer by fusing network representation and deep learning. The application accurately identifies disease core driver genes through single-cell transcriptome differential analysis and protein-protein interaction network topology centrality index, simultaneously extracts the structural characteristics of traditional Chinese medicine ingredients by graph isomorphism network, constructs traditional Chinese medicine-compound isomorphism network, and extracts the isomorphism topological representation of traditional Chinese medicine by network representation algorithm. The one-dimensional convolutional neural network is used to deeply mine the protein sequence nature semantic features of disease gene sequence, so that the protein features and the end dimension of traditional Chinese medicine are aligned. Finally, the cross-modal joint vector is constructed through feature splicing, and the intervention probability output by the full connection neural network is used to realize the scoring of the candidate drug. The application establishes a direct and quantitative connection between the traditional Chinese medicine intervention space and the single-cell pathological mechanism, realizes the end-to-end evaluation of the anti-colorectal cancer efficacy of traditional Chinese medicine, and can be applied to the fields of traditional Chinese medicine screening and individualized drug administration scheme making.

This invention discloses a method and system for detecting icing thickness on power transmission lines based on a large visual model. It employs a higher-performance segmentation model to process key frames in power transmission monitoring videos, accurately extracting the icing coverage area. Simultaneously, through temporal comparative analysis of different frame images, combined with the dynamic changes in icing features between frames, it achieves accurate real-time estimation of icing thickness. This method not only overcomes the shortcomings of shallow feature extraction in small models but also significantly improves the accuracy and real-time performance of icing detection by relying on video temporal data, making it more suitable for the operation and maintenance needs of power transmission systems. Addressing the real-time and accuracy requirements of power transmission line icing detection, this invention constructs a three-level linkage technical architecture of "target detection - visual segmentation - frame differential analysis," relying on real-time video streams acquired by video monitoring devices already deployed on power transmission lines to achieve fully automated detection of icing location, accurate segmentation, and thickness estimation.

This invention relates to the field of energy management technology, specifically to an artificial intelligence-based new energy energy-saving scheduling and carbon control system, including an output acquisition module, a phase correction module, a path mapping module, a scheduling and orchestration module, and a carbon scheduling generation module. In this invention, differential analysis of multi-source output and load operating status is used to extract change nodes, establishing a correspondence between output behavior and load actions on the time axis. A consistent adjustment range is constructed by combining the adjustment command interval and the output response offset relationship. Simultaneously, a segmented carbon path structure is formed based on the relationship between equipment operating status and energy flow, coupling and characterizing energy type and flow process. Energy flow direction and continuous characteristics are divided according to time intervals, forming an execution sequence matching the carbon path and associating emission coefficients with energy flow direction. This constructs a carbon emission correlation covering the entire adjustment process, strengthening the correlation between control behavior and emission results, and improving the refinement of scheduling decisions.

This invention relates to the field of photoelectric physiological sensing technology and discloses a non-invasive health monitoring system based on photoplethysmography pulse wave and multi-sensor fusion, comprising: a multi-wavelength photoelectric sensing module, a thermoelectric sensing module, a motion vector acquisition module, and a physiological feature restoration unit. This unit identifies the pressure state based on the rate of change of acceleration data, extracts a reference green light recovery curve and calculates the blood flow filling time constant, determines vascular compliance parameters using local body temperature and performs nonlinear correction based on the filling time constant, generates motion artifact components with tissue pressure hysteresis characteristics by combining acceleration data, and restores the physiological pulse wave signal from the original signal by differential analysis. This invention establishes a viscoelastic model by capturing the physiological refill process, achieving physical decoupling between physiological signals and motion noise, and ensuring the stability of the monitoring benchmark under dynamic conditions.