26results about How to "Eliminate noise interference" patented technology

The application discloses a comprehensive linkage alarm system for fire hazards and structure instability of timber buildings, and relates to the technical field of timber safety monitoring, which comprises the following modules: a multi-dimensional heterogeneous signal acquisition module for collecting data and constructing real-time state monitoring data sets; a structure spectrum benchmark modeling module for constructing a thermal-rigidity attenuation theoretical model; a thermal-frequency correlation analysis module for generating a thermal-frequency deviation degree index; a structure state determination module for identifying timber pyrolysis states and structure instability risks; a cascade response module for executing corresponding fixed-point spraying control, evacuation path planning and alarm output; and a self-adaptive feedback correction module for warehousing data and correcting thermal aging parameters. The application establishes a correlation model of thermal and rigidity by fusing thermal imaging and vibration data, effectively eliminates environmental wind load noise, realizes accurate identification of internal hazards of timber buildings, and improves the accuracy of building monitoring and the efficiency of emergency response.

The application belongs to the technical field of power systems and their automation, and specifically discloses a method for stabilizing electromagnetic oscillation of a photovoltaic grid-connected system considering dynamic characteristics of a converter. According to the application, a target current absorbed by a target inverter from a DC side in the photovoltaic grid-connected system is calculated based on a target signal processing device; a target DC bus voltage value is calculated based on a DC bus voltage equation; a difference between a DC bus voltage reference value and the target DC bus voltage value is calculated, and the action of the target converter is controlled according to the voltage difference. In the above manner, the target current is calculated by taking the filtered DC bus voltage measurement value as the divisor and the power as the dividend, so that the target current no longer sharply increases with the drop of the actual voltage under high-frequency dynamics, and the oscillation between the circuit and the constant power load is suppressed by introducing positive damping in the control loop, thereby significantly improving the stability of the photovoltaic grid-connected system and reducing the cost of stabilizing electromagnetic oscillation.

The invention relates to the field of intelligent power grids and intelligent fusion terminals, in particular to a transformer health monitoring system based on an intelligent fusion terminal TTU, and the system comprises a waveform data collection module which is used for collecting the waveform data of a transformer; the environment data acquisition module is used for acquiring operation environment data of the transformer; the denoising and reconstruction module is used for carrying out denoising elimination and waveform reconstruction on the waveform data of the transformer; the feature extraction module is used for performing feature extraction, enhancement and fusion on the preprocessed waveform data and the operation environment data of the transformer to obtain a fusion feature vector; the state analysis module is used for inputting the fusion feature vector into a mixed regression model for health state prediction to obtain a transformer health index prediction value; and the TTU decision module is used for matching the control decision scheme and controlling the working state of the transformer. According to the invention, the accuracy and reliability of transformer health prediction and the reliability of overall operation of a power grid can be improved.

The application discloses a microseismic event positioning method based on a microseismic signal, and is characterized by comprising the following steps: S1, collecting the microseismic signal by using a microseismic collection device; the microseismic collection device comprises a plurality of three-component microseismic detectors which are arranged at equal intervals in a long drill hole and are used for collecting the microseismic signal generated by a microseismic event in a deep coal seam; the P wave and the S wave of the microseismic signal are collected simultaneously by using the microseismic collection device, and the time difference between the P wave and the S wave is used to directly provide the absolute distance from each receiver to the seismic source, so that the positioning process is converted into a spherical surface intersection problem with each receiver as a spherical center and with a calculated distance as a radius. The method not only eliminates systematic errors caused by the uncertainty of the time of earthquake occurrence, but also provides a large amount of redundant data for cross verification, so that the accuracy and reliability of the positioning result are greatly improved. The accuracy and comprehensiveness of signal collection are improved.

This invention discloses a method and apparatus for accurate bridge damage identification and location based on multi-source data fusion. The method includes: performing spatiotemporal registration and data fusion on the time-series data of the bridge's global planar deformation obtained by ground-based synthetic aperture radar and the time-series data of the bridge's point-like three-dimensional deformation obtained by the BeiDou satellite navigation system to obtain a fused deformation field of the entire bridge area; calculating multiple damage characteristic indicators reflecting different deformation physical mechanisms of the bridge based on the fused deformation field, and determining the corresponding preliminary damage level based on the damage judgment threshold of each damage characteristic indicator; using the preliminary damage levels of each damage characteristic indicator as independent evidence sources, performing evidence fusion to resolve indicator conflicts, and determining the bridge damage location and damage level based on the fusion result. This method can effectively improve the accuracy and reliability of bridge damage identification and location, providing a scientific basis for bridge operation and maintenance decisions.

The invention discloses an energy storage energy management method, device and system and a medium, the energy storage energy management system comprises a local energy management system, a local energy management system and a local energy system, the local energy management system obtains operation data of each device in the local energy system, the model prediction precision of a locally deployed energy prediction model is improved; based on the data type of the operation fusion data, performing data fusion processing on all the operation data to obtain the operation fusion data; calling an energy prediction model to perform energy prediction on the operation fusion data to obtain a strategy prediction result; when the model prediction precision reaches the target precision, obtaining a corresponding local pre-stored strategy according to the operation fusion data, and adjusting the local pre-stored strategy according to a strategy prediction result to obtain a target control strategy; according to the invention, the accuracy of energy management and control can be improved, so that the reliability of an energy storage energy management system is improved.

The application relates to the technical field of artificial intelligence, and discloses a speech synthesis method and device, computer equipment and a storage medium, which comprise the following steps: obtaining reference speech data and speech synthesis text; performing emotion extraction on the reference speech data to obtain reference emotion embedding; fusing the reference speech data, the speech synthesis text and preset noise features to obtain original fusion features; determining a dynamic emotion injection coefficient based on a preset denoising time step parameter; performing denoising processing on the original fusion features based on the dynamic emotion injection coefficient and the reference emotion embedding to obtain target synthesis spectral features; and performing audio conversion based on the target synthesis spectral features to obtain target speech data. The application can be applied to the speech synthesis scene of financial technology and medical health, and improves the emotion expression accuracy of speech synthesis.

This disclosure provides an image detection method, apparatus, electronic device, and storage medium. The image detection method includes: acquiring a multiplex immunofluorescence image to be tested, wherein the multiplex immunofluorescence image is imaging data after multiplex immunofluorescence staining of a tissue sample to be tested; performing tissue segmentation based on the multiplex immunofluorescence image to obtain a tissue region image; and performing image sharpness detection on the tissue region image to obtain a detection result. The image detection method provided by this application can effectively eliminate background noise interference in multiplex immunofluorescence images caused by staining at different locations in the image, greatly reducing the proportion of background in the input image for subsequent image detection, and improving the accuracy and reliability of the image detection results.

This invention discloses an algorithm for removing high-power noise interference based on vortex waves, belonging to the field of signal processing technology. In this invention, the transmitted signal is a vortex wave signal. At the receiving end, an antenna array (number of array elements N≥2) receives the signal containing high-power noise interference. The received mixed signal is then subjected to vortex wave-based noise reduction processing. The vortex wave noise reduction principle is as follows: The average amplitude or the average of the maximum and minimum amplitude values ​​and the average phase of the electric field of N echo signals are taken. Then, the average amplitude and phase values ​​are combined to form a noise reduction electric field. Alternatively, one echo signal from the N echo signals can be arbitrarily selected, and the noise reduction electric field is subtracted from that echo signal. Compared with existing plane wave-based signal processing algorithms, this invention can extract vortex wave signals under low signal-to-noise ratio conditions, which is beneficial for improving the anti-interference capability in the field of signal processing and has significant practical application value.

The invention discloses a vehicle mobile phone volume change track detection method, device and equipment and a storage medium. According to the method, a Python automation script is used for controlling terminal equipment to execute a sound detection process, a target sound source is automatically triggered, a volume gear is adjusted, a sound scene is switched, and an operation environment is provided for sound data collection; according to an operation environment, a browser sound collection interface is synchronously called to capture physical parameters of loudspeaker sound of terminal equipment in real time, original collection data are obtained, the original collection data are calibrated and filtered, standardized sound data are generated, and the standardized sound data are displayed in real time through a visual interface. And after the detection is completed, the full-amount detection data is automatically packaged and stored according to the preset format, so that the full-automatic operation of the volume change track detection of the vehicle-mounted terminal and the mobile phone can be realized, the detection efficiency is remarkably improved, the subjective error of manual operation is eliminated, the environmental noise interference is effectively overcome, and the objectivity and accuracy of the detection data are ensured.

The application provides an artificial intelligence-based call center data analysis method, and relates to the field of call center data processing.The method collects real-time call audio and pre-processes it, extracts mel-frequency spectrum features and text data with timestamps;uses an analysis model integrating acoustic layer, text layer, cross-modal fusion layer and multi-task analysis head layer to obtain multi-dimensional analysis results such as emotional state, service compliance, service quality score and customer intent.In model training, an adversarial gradient reversal layer and an acoustic environment discriminator are introduced to eliminate environmental noise interference, and a timing risk encoder and a contrast learning are used to enhance the risk evolution capture capability.Based on the analysis results, a comprehensive risk index is calculated, an intervention strategy is matched and real-time pushing is performed, and a comprehensive analysis report is generated after the call is completed.The application realizes cross-environment adaptive purification and multi-modal fusion analysis, effectively improving the service quality monitoring and risk early warning efficiency of the call center.

The application provides a cross-environment software consistency determination method and device based on multi-dimensional dependency analysis, which comprises the following steps: collecting the build deployment information of the target software from the build deployment tool and the code controlled library; establishing a benchmark configuration library containing the build deployment information for each environment respectively; obtaining the real-time hash value of the target software binary file under different test environments, and comparing the real-time hash value with the file-level hash value in the benchmark configuration library to obtain a first comparison result; in response to the first comparison result passing, comparing whether the source code versions of the target software and all its dependencies under different test environments are consistent, and whether the dependency relationship topologies are isomorphic based on the source code version information and the dependency fingerprint to obtain a second comparison result; in response to the second comparison result passing, verifying whether the environment difference is within a preset compliance range; and in response to the environment difference being within the preset compliance range, determining that the target software under the cross-environment is consistent.

The invention discloses a multi-modal data fusion-based power equipment monitoring method and monitoring system. The method comprises the following steps of: acquiring image information and sensor data, and preprocessing the image information and the sensor data; image features of the preprocessed image data and sensor features of the sensor data are extracted, weighted fusion and PCA dimension reduction are carried out on the image features, the sensor features and cross-modal attention fusion features, and a final fusion feature vector is obtained, the fusion weight of the image feature and the sensor feature is dynamically adjusted according to the equipment operation state feature value, and the cross-modal attention fusion feature is the cross-modal attention fusion feature of the image feature and the sensor feature; and obtaining an equipment state and a risk level through forward propagation based on the final fusion feature vector. According to the monitoring method and the monitoring system, through combination of image processing and sensor data analysis, the accuracy of state monitoring of the power equipment is improved.

The application relates to the technical field of new energy automobile battery safety management, in particular to a new energy automobile battery monomer short circuit detection system and method, the system collects target battery signals and battery physical and chemical data through a battery data acquisition module, filters, normalizes and extracts features of the target battery signals through a battery signal processing module, dynamically adjusts a short circuit feature self-adaptive threshold according to real-time working condition quantitative values through a monomer short circuit monitoring module, predicts a short circuit development trend and provides an intervention window period through a short circuit trend prediction module, and monitors an electric-power coupling value and dynamically adjusts and regulates a regulation strategy through a risk window regulation module, so that early and accurate detection and graded early warning of battery monomer short circuits are realized, and the accuracy of short circuit detection and the timeliness of graded regulation are improved.

The invention provides a stirring device of a shield tunneling machine, which solves the problem that accumulated materials are difficult to scrape off through a stirring blade and accumulate for a long time due to the fact that the stirring blade of the existing stirring machine adopts a fixed-length structure and a gap exists between the stirring blade and the inner wall of a tank body, and can be widely applied to the field of stirring machines. The device specifically comprises a tank body, a stirring shaft is arranged in the tank body, a plurality of fixing pipes are distributed on the surface of the stirring shaft, stirring blades are connected into the fixing pipes, first blades are arranged at the top ends of the stirring blades, second blades are connected to the surfaces of the first blades, the shapes of the second blades are matched with the inner wall of the tank body, and the ends of the second blades can be attached to the inner wall of the tank body; the ends of the second blades protrude out of the ends of the first blades, so that when the ends of the second blades are attached to the inner wall of the tank body, only the second blades make contact with the inner wall of the tank body.

The invention relates to the technical field of multi-target tracking and azimuth angle high-precision measurement, and discloses a target point azimuth angle high-precision measurement method in large-size antenna measurement, which comprises the following steps of: firstly, acquiring an image of a measured target which comprises a plurality of preset mark points, and then processing the image, extracting the physical coordinates of all the target points contained in the image, the physical coordinates of all the target points including the coordinates of the plurality of preset mark points and the coordinates of the interference points; and matching the preset model of the measured target with the physical coordinates of all the target points. By constructing a preset model of a measured target and matching the model with physical coordinates (including mark points and interference points) of all extracted target points, coordinates of a plurality of preset mark points can be accurately identified and separated from a complex image background; and the accuracy of target point identification and the anti-interference capability of the whole measurement method are obviously improved.

The application relates to the technical field of semiconductor equipment monitoring, in particular to a guide rail cleanliness monitoring system and method for semiconductor equipment; the system comprises a signal acquisition module, a signal processing module, a pollution identification module and an evaluation decision module; the signal processing module performs signal analysis on the collected electrostatic induction signals and acoustic emission signals, and performs event detection in combination with motion state information to obtain a characteristic event list and a pollution index vector; the pollution identification module performs space-time correlation analysis on the characteristic event list to obtain internal wear sources, and identifies external deposition sources in combination with the pollution index vector; an excitation signal is applied to the external deposition sources, and a silent pollution layer is identified according to a response result; the evaluation decision module generates a comprehensive evaluation result of the guide rail cleanliness state based on the internal wear sources, the external deposition sources and the silent pollution layer; the application realizes online monitoring of guide rail cleanliness and accurate identification of pollution sources, and improves the reliability of semiconductor equipment operation.

The invention relates to the technical field of power equipment monitoring, in particular to a transformer state online monitoring and fault early warning device and method. According to the technical scheme, the transformer state online monitoring and fault early warning device comprises an early warning device body, a control unit, an environment sensor, a wiring port, a current sensor, a wireless communication mechanism, a buzzer and a protection plate, the control unit is arranged at the bottom end of one side of the early warning device body, and the environment sensor is arranged on one side of the control unit; the environmental sensor is introduced to realize real-time monitoring of the humidity of the operating environment, a collaborative analysis mechanism of environmental data and electrical data is established, the environmental sensor continuously acquires the humidity value of the operating environment of the transformer, and the control unit performs digital filtering processing on the humidity data to eliminate environmental noise interference. After the standardized humidity value is generated, the standardized humidity value and the current data jointly participate in fault diagnosis, and deep correlation between environmental factors and equipment states is achieved.

This invention discloses a video transmission and reception frequency alignment method, apparatus, and system, belonging to the field of video transmission and display control. The method includes: a transmitting end sampling and recording a first timestamp Tn at preset intervals and sending a trigger message; a receiving end receiving the message and latching a second timestamp Rn via hardware; calculating a first difference Dn = Rn - Tn and a difference value ΔDn = Dn - D(n-1); filtering Dn and / or ΔDn; and fine-tuning the local output video clock frequency based on the filtering result to make Dn tend to be constant and ΔDn approach zero, thus achieving frequency alignment. The apparatus includes a transmitting module, a receiving module, a calculation module, a filtering module, a control module, and a judgment module. The system includes a transmitting end and a receiving end. This invention does not require a shared absolute reference clock; the paired timestamp correspondence is clear; it is insensitive to fixed link delays; and it balances smooth video display and sub-frame-level low latency.

The present application relates to the technical field of image processing, and more particularly to a method and system for agricultural monitoring and decision-making based on multi-source perception and image analysis. The method first collects multi-source images of farmland, including foggy day crop tilt images, and obtains high-quality images through registration fusion and smoothing denoising processing. Potential boundary points are marked using pixel gradient calculation, and a refined boundary image is generated by combining edge detection algorithms to determine high-precision boundary positions through gradient feature analysis. Based on the boundary positions, the pixel gradient direction distribution is integrated to form a set of crop linear structures, and the stem tilt trend is tracked through linear fitting and direction statistics to generate a stem tilt direction trend graph. The water-sensitive spectral features in the hyperspectral image are extracted, and the morphological information of the stem tilt trend is fused to input a pre-trained soil moisture evaluation model to generate a soil moisture deviation evaluation graph representing the field water distribution condition. This technical solution provides data-driven decision support for agricultural precision irrigation and resource optimization.

The present application relates to the technical field of electric digital data processing, in particular to a battery thermal runaway early identification system based on multi-modal perception fusion and charging process. The system obtains electric signals and multi-source sensing signals in the charging process. A causal alignment processor extracts the voltage rate of change extreme point caused by the charging step in the electric signal as a time reference anchor point. Using the anchor point as a constraint boundary, the low-frequency sampling waveform is mapped to the high-frequency time axis using the dynamic time warping algorithm to generate a causal alignment multi-modal joint time sequence matrix. A thermal runaway evolution graph calculation model maps the matrix eigenvalue to a graph node, maps the mutual information of adjacent time stamps and cross-modal features to a graph edge weight, and extracts a subgraph structure with abnormally aggregated weights through graph convolution operation. When the edge weight cumulative value of the subgraph structure exceeds the baseline threshold, an identification signal is output. The present application eliminates the time offset interference caused by the sampling difference and physical time delay of the sensor, and extracts the cross-modal causal correlation features.

The invention provides a method for quickly identifying hot spots of a divertor of a fusion device under a cyclic thermal load. The method comprises the following steps of: recording surface temperature distribution and an average value once by using an infrared thermal imager every N / 5 cycles; calculating absorption heat flux density through coolant inlet and outlet temperature difference, and screening effective heat load data; monitoring inlet flow, temperature and pressure; setting a relative percentage threshold as a standard acceptance factor AC; the average temperature of each tungsten block is firstly compared integrally, then local highest temperature comparison is carried out on abnormal blocks, and finally global analysis is carried out. The relative temperature rise acceptance factor is introduced based on the minimum average temperature, and the judgment statistical significance is improved; dynamically calculating the heat flux density to ensure the loading effectiveness; the method is suitable for divertor delivery testing and fusion reactor in-situ monitoring, rapid and automatic hot spot identification is achieved, and the testing efficiency and accuracy are improved.

This application relates to an intelligent welding tracking method, system, device, and medium for alloy chain plates, belonging to the field of intelligent welding technology. The method includes: acquiring point cloud data and image frame data of the area to be repaired; converting the point cloud data to a base coordinate system to obtain clean point cloud data, and identifying all points below the unworn surface with a difference exceeding a set height threshold to form the point cloud contour of the worn area; performing edge detection on the image frame data to obtain edge pixels containing the wear boundary and crack lines, and back-projecting them to the base coordinate system to generate a projection point set; fusing the projection point set and the point cloud contour according to a fusion distance threshold, and applying B-spline curve fitting to generate the contour baseline of the area to be repaired; selecting suspected crack edge segments from the edge pixels and mapping them back to the contour baseline while recording crack parameters; generating a welding path based on the contour baseline of the area to be repaired and the crack parameters, thereby achieving automatic welding and improving the stability of the alloy chain plate repair quality.

The application provides a mine underground safety prevention and control method and device based on a vehicle-mounted device, equipment and a storage medium, and relates to the technical field of mine safety management and control. Through local preprocessing of the vehicle-mounted end to eliminate environmental noise interference, real-time fusion and global optimization of data are realized in combination with an edge-cloud collaborative architecture, and a network interruption autonomous mechanism and a dynamic risk assessment control algorithm are innovatively introduced, effectively solving the problems of sensing failure and network interruption safety loss of control caused by extreme working conditions underground, solving the reliability problem of environmental sensing data in an extreme environment underground, realizing local autonomous safety control when the network is interrupted, and improving the real-time performance and reliability of mine underground safety prevention and control.

The invention relates to the technical field of electric power industry, and discloses a BI data technology-based power grid cost review auxiliary method, system, device and medium, and the method comprises the steps: integrating multi-source engineering data, and carrying out cleaning and standardization processing to construct a multi-dimensional data cube; calculating a list item reference unit price based on historical data and predicting a project cost, and comparing the project cost with a submitted value to realize quantitative analysis of a project level and list level deviation; constructing a risk comprehensive scoring model in combination with the historical abnormal rate and the current deviation characteristics, dividing early warning levels, and screening key review list items; and realizing risk visual display based on a BI platform, and automatically generating an auxiliary report containing review key points and suggestions. According to the method, full-process digital review from data integration, intelligent analysis, risk early warning to decision support is realized, and the review efficiency, the anomaly recognition accuracy and the resource allocation rationality are effectively improved.

This invention discloses a method, apparatus, device, storage medium, and system for verifying the layup orientation of carbon fiber composite materials based on polarization imaging. The method acquires light intensity images from multiple polarization angles, reconstructs the surface microscopic normal vector field, employs von Mises-Fischer probability distribution statistical modeling, utilizes a dual neural network model to map the roughness sequence, constructs an angular domain power spectral density function after data correction, extracts the actual layup orientation angle, and completes quality grading verification. This invention achieves high-precision, non-destructive, and automated detection of the layup state of carbon fiber composite materials, effectively solving the problems of destructive, inefficient, and insufficient precision in existing detection technologies, and meeting the quality control needs of high-end equipment manufacturing and industrial mass production of carbon fiber composite components.