30 results about "General algorithm" patented technology

The embodiment of the invention discloses a waveform identification and correction method and system for winding deformation diagnosis, and relates to the technical field of transformer winding monitoring, and the method comprises the steps: synchronously collecting voltage and current signals, and carrying out the preprocessing; performing waveform disturbance type identification on the preprocessed signal, and dividing the disturbance type into harmonic dominant disturbance, voltage sag disturbance or compound disturbance, and harmonic dominant disturbance, voltage sag disturbance or compound disturbance; dynamically selecting a corresponding waveform processing path according to a waveform disturbance type identification result; processing the signal by adopting the selected waveform processing path so as to correct and reconstruct the signal and obtain compensated voltage and current waveform data; by constructing a disturbance identification and path adaptive processing mechanism, the problems of processing lag and compensation distortion of a general algorithm during dynamic signal disturbance are solved, the identification precision is improved, and real-time error correction is realized.

The present invention relates to the technical field of motor design, and provides a method for calculating a leakage magnetic coefficient of a closed slot in a rotor of an asynchronous motor. According to the calculation method provided by the present invention, by means of diverting the magnetic flux of rotor teeth in combination with a relative magnetic permeability calculated from a magnetization curve, a closed slot is ultimately transformed equivalently into a semi-closed slot, thereby unifying the semi-closed-slot and closed-slot algorithms and improving the adaptability of the closed-slot algorithm. The present method provides a general algorithm for closed-slot parameters, takes into account the edge effects caused by punching and shearing of cold-rolled silicon steel sheets, and is applicable to closed slots of various rotor slot types. Compared to finite-element methods, the present method has a short calculation cycle and is suitable for programmatic processing.

The invention discloses an algorithm arrangement method and a related device, and the algorithm arrangement method comprises the steps: receiving a plurality of nodes selected by a user from a node library, and receiving a node sequence set by the user for the plurality of nodes, the plurality of nodes comprise a plurality of general algorithm nodes and a plurality of scene adaptive nodes, a scene adaptation node is associated with at least one general algorithm node; according to the node sequence, the multiple nodes are combined to form a target algorithm, the target algorithm is used for data classification of a target scene, in the data classification process, the universal algorithm nodes are used for data classification of the to-be-classified data, and a first classification result is obtained; and the scene adaptation node is used for correcting the first classification result obtained by the associated general algorithm node to obtain a second classification result. According to the scheme, the data classification effect of the target algorithm in the target scene can be improved.

This invention belongs to the field of power monitoring technology and discloses a brightness balancing and regional smoothing fusion algorithm for power monitoring images. The method includes: image denoising, using mean filtering and median filtering to smooth the image; image registration, performing optimal matching on two or more images obtained from different sensors, at different times, or from different angles; and image region fusion, smoothly fusing images with different angles and brightness contrasts. This method is applicable to panoramic video monitoring systems. Designed specifically for the complex lighting and high electromagnetic interference characteristics of power operation sites, this invention effectively eliminates on-site noise through adaptive preprocessing, and achieves a balance between global brightness uniformity and local detail preservation through layered brightness balancing, solving the problems of incomplete brightness balancing and detail loss in power scenarios encountered by general algorithms.

The invention relates to the technical field of transmission network equipment maintenance, in particular to a PTNSPNOTNSOTN equipment-oriented performance anomaly root cause positioning method and system, and the method comprises the steps: constructing an equipment exclusive data collection system, constructing a hierarchical collection model based on equipment types, collecting and preprocessing data of an equipment layer and a network layer, and constructing an equipment state matrix containing equipment type labels; designing an inference model adaptive to the equipment type, and outputting a sorted root cause list through first-layer equipment type-exception type mapping, second-layer hierarchical root cause candidate generation and third-layer cross-layer root cause fusion; the method has the beneficial effects that a data acquisition and reasoning module is designed specially for exclusive characteristics and protocols of PTN / SPN / OTN / SOTN equipment, the problem of poor adaptability of a general algorithm on the equipment is solved, and the characteristic coverage rate is increased by 70%. Key indexes and abnormal features of various devices can be accurately captured, and a comprehensive and accurate data basis is provided for root cause positioning.

The invention discloses a digital twinning system generation method, computer equipment and a storage medium, and belongs to the technical field of digital twinning and information physical systems. The method comprises the following steps: S1, carrying out general algorithm operator packaging; s2, operator modeling combination is carried out; and S3, performing intelligent agent enabling on the model. According to the invention, high modularization and reusability are realized, and the development cost and repeated labor are greatly reduced; the expandability and maintainability of the digital twin system are obviously improved, and the system is endowed with the ability of sustainable evolution; a construction path and a cooperation framework of the intelligent system are clarified, and the team cooperation efficiency is remarkably improved; intelligent layering from calculation to autonomous decision making is achieved, and cognitive intelligence is introduced to solve complex problems.

The application discloses an image generation method and device based on a diffusion model, constructs a conditional vector generation module containing detail preservation, an attention-based diffusion model and a secondary distillation iteration framework, accelerates the running speed of the diffusion model, minimizes the influence on the quality of generated images, achieves a better balance between reducing model running loss and improving image quality, introduces a conditional attention mechanism into the diffusion model, encodes the conditional vector by using the multimodal adaptability of the self-attention mechanism, and simultaneously enhances the capture of image details, and can be used as a general algorithm framework in the field of image generation.

This application discloses a system for building an action recognition algorithm platform based on multi-algorithm fusion, relating to the field of machine learning technology. The system includes: a model inference layer, an application algorithm layer, and a scenario strategy layer. The model inference layer receives multimodal input data, performs model inference processing on the multimodal input data through multiple parallel inference links, and outputs standardized inference results with a unified timestamp. The application algorithm layer aligns multi-source data based on a unified timestamp and executes multiple general algorithm functions in parallel, outputting standardized algorithm function results with a unified timestamp. The scenario strategy layer receives the standardized algorithm function results with a unified timestamp from the application algorithm layer, combines and calls corresponding general action recognition algorithm functions according to the business requirements of the target medical assessment scenario, executes action assessment logic matching the target medical assessment scenario, and finally outputs the action assessment results for that medical assessment scenario.

The invention provides an AI video monitoring platform based on an RISC-V framework. The AI video monitoring platform comprises eight levels of frameworks including an equipment perception layer, an edge computing layer, a data access layer, a data processing layer, a data storage layer, an algorithm engine layer, a platform application layer and an industry application layer. The invention belongs to the technical field of monitoring platforms, particularly relates to an AI video monitoring platform of an RISC-V architecture, and aims to at least solve the problems that traditional security depends on manual screen staring and inspection, abnormal event recognition lags behind and the missing report rate is high, and the existing system is high in authorization cost, insufficient in edge scene energy efficiency ratio and poor in adaptability. The risks that the data transmission delay is high, real-time early warning cannot be supported and privacy leakage exists in original video cloud storage due to a front end-cloud framework are avoided; the method avoids the phenomena that different brand devices and platform protocols are not uniform, the compatibility is poor, the deployment and maintenance cost is high, and the expansion capability is weak, and overcomes the defects that a general algorithm is insufficient in precision, lacks an industry exclusive algorithm, and cannot cover the multi-scene safety protection requirement.

The invention discloses a 4K ultra-high-definition real-time cloud stream pushing and rendering method and system, and the method comprises the steps: training an over-fitting super-resolution model for each virtual scene at a cloud end, transmitting the over-fitting super-resolution model to a client, only needing to transmit a low-quality image, greatly reducing the bandwidth demand and delay, and achieving the real-time cloud stream pushing and rendering. And meanwhile, a lightweight super-resolution model is utilized to process a low-quality image at a client to generate a 4K output image, so that the performance bottleneck of low-configuration equipment is broken through, and real-time rendering of 4K ultra-high-definition image quality can be realized in a resource-limited environment. Besides, through mechanisms such as multi-scale gradient loss and dynamic space attention, scene adaptability and detail performance are enhanced, the problem of detail blurring of a general algorithm is avoided, and therefore the visual effect and processing efficiency far better than those of a traditional method are provided in a specific virtual scene. And the obvious improvement is brought to the improvement of the user experience in the fields of digital twinning, analogue simulation, cloud games and the like.

The invention discloses an intelligent early warning system for construction safety along a high-speed railway, and the system comprises a front-end sensing layer which is disposed in a construction area along the high-speed railway, comprises a plurality of high-definition long-focus full-color cameras, and is used for collecting the real-time video data of the construction area; the network transmission layer is connected with the front-end sensing layer and is used for transmitting the real-time video data to a rear end; and the AI intelligent analysis layer is connected with the network transmission layer and comprises a computing power server and an AI analysis model deployed on the computing power server. According to the method, an AI large model technology is applied, and an L2-level AI large model subjected to building industry pre-training and adjacent business line scene deep customization training is applied to the field of high-speed railway construction safety for the first time. The hierarchical training mode not only ensures that the model has a wide visual understanding basis, but also enables the model to achieve extremely high accuracy and robustness in a specific task, and overcomes the problem that a general algorithm is poor in recognition effect in a complex railway scene.

This invention discloses a general algorithm verification platform and its algorithm registration and invocation method. The platform provides an open-source algorithm plugin framework template, which includes multiple languages. Users integrate algorithms using this framework and fill in the processing results according to the network interface protocol requirements for algorithm-platform interaction. User algorithms do not need to consider network communication. The interface between the algorithm and the platform uses TCP / IP communication. The uploaded algorithm acts as a TCP client, actively initiated by the platform, and receives connection information from the TCP server, as well as initialization parameters. The algorithm continuously attempts to connect with the platform. When a connection is successful, it automatically sends process ID information, waits for platform data input, the platform transmits data to the algorithm, and the algorithm feeds back the processing results to the platform. This invention facilitates the verification and integration of different algorithms, thereby minimizing user development costs.

The invention discloses a general algorithm modeling method, system and platform, electronic equipment and a storage medium, and belongs to the technical field of computer software, and the method comprises the steps: analyzing a predefined protocol file in an algorithm package, and obtaining the configuration information of an algorithm; dynamically generating a visual configuration interface based on the configuration information to receive configuration input of a user; according to the configuration information and the configuration input, automatically constructing a mirror image which can be operated by an algorithm, and allocating resources during operation; and loading and executing the mirror image in the runtime resource, and responding to a data and parameter request in an algorithm execution process through an agent component embedded in the algorithm so as to complete algorithm operation. According to the method, decoupling of the algorithm and an engineering system is achieved through the protocol file, full-process automation and standardization of the algorithm from development, configuration, release to operation are achieved through the containerization and dynamic scheduling technology, and the algorithm delivery efficiency and reusability are remarkably improved.

This invention relates to the field of power data storage and transmission technology, and discloses a method, transmission system, device, and electronic equipment for compressing electricity consumption data. The method includes: acquiring raw electricity consumption data; generating an electricity consumption data matrix based on the raw electricity consumption data; determining whether the data in the electricity consumption data matrix is ​​valid and generating matrix markers; filtering valid data based on the matrix markers and compressing the valid data using a compression algorithm; and generating compressed data blocks by combining the matrix markers. This invention transforms domain knowledge into compression rules, overcomes the shortcomings of general algorithms in utilizing the characteristics of structured data, significantly reduces the transmission bandwidth and storage cost of massive amounts of electricity consumption data, and simultaneously meets the real-time requirements of high-frequency acquisition scenarios.

The invention relates to the technical field of engineering digital management and intelligent scheduling, in particular to a dynamic allocation method and device for engineering resources. The method comprises the following steps: acquiring multi-source heterogeneous physical data of a full life cycle of a project, and preprocessing the multi-source heterogeneous physical data; based on the preprocessed physical data, mining a nonlinear relationship between environmental physical parameters and resource consumption by using a gradient boosting decision tree, and predicting a physical resource demand correction coefficient and a working condition risk level of each construction node; and outputting an optimal resource allocation vector on the premise that physical constraints are met by adopting a He-Marx optimization algorithm, a physical resource demand correction coefficient and a working condition risk level and aiming at minimizing physical supply and demand deviation, system operation energy consumption and safety out-of-limit penalty. The method can solve the problems that in the prior art, static planning is difficult to match with real-time changing field requirements, and a general algorithm is difficult to process complex physical constraints, so that resource distribution is not uniform.

The invention relates to the field of PDF (Portable Document Format) electronic signatures, in particular to a double-algorithm signature and verification method of a PDF document. The method comprises the following steps: a signature process: calculating to obtain a first digest value and a second digest value; generating a second signature structure data object based on the second digest value; a first signature structure data object is generated based on the first digest value and a second signature structure data object, and the second signature structure data object serves as an attribute to be placed in an authentication attribute set of the first signature structure data object; a verification process: extracting a first signature structure data object from the PDF document, and verifying the validity of the first digital signature and the second digital signature in sequence; and respectively extracting the first digest value and the second digest value, and comparing the first digest value and the second digest value with corresponding reference digest values to confirm the integrity of the document. According to the method, on the premise that PDF related standard specifications are completely followed and compatibility is ensured, collaborative signature protection of the international general algorithm and the national cryptographic algorithm of the same PDF document is achieved.

The present invention provides a model predictive control strategy for an optimal switching sequence of MMC based on output voltage space vector fitting. The strategy calculates the coordinates of each candidate vector in the g-h coordinate system from the vector prediction value according to the discrete prediction model of the output voltage space vector of the three-phase MMC AC side; obtains the optimal switching sequence by screening the redundant switching states of the candidate vectors; constructs a value function of the sum of squared output current errors based on the current change gradient and obtains the optimal action time expression by taking the partial derivative thereof; calculates the optimal duty cycle of the bridge arm and the number of submodules turned on according to the bridge arm unbalanced voltage drop prediction model; and achieves submodule capacitor voltage balance through a sorting algorithm. The benefit of the present invention is that compared with the traditional model predictive control strategy, the strategy has a small computational burden, effectively reduces the output current harmonic content, and has a more concentrated harmonic spectrum distribution. Taking MMC as an example, a general algorithm for screening the optimal switching sequence of a multi-level converter is designed, which can be applied to any level and has a constant computational load.

The invention discloses a specific field text file custom dictionary auxiliary Huffman compression method and system, and belongs to the technical field of data compression. The method comprises the following steps: receiving and verifying a phrase-short code mapping dictionary customized by a user, and ensuring that the length of a short code is smaller than that of an original phrase and ambiguity is not generated; sorting the dictionaries in a descending order according to the phrase length to form a replacement priority queue; performing phrase-to-short code replacement on the text according to a priority sequence by adopting a block processing strategy; carrying out character frequency statistics on the replaced text and constructing a Huffman tree; generating a code based on a Huffman tree, and converting the text into a bit stream; and generating a structured compressed file containing the metadata area, the dictionary information area and the compressed data area. The system comprises a dictionary management module, a text replacement module, a Huffman coding module and a file storage module. According to the method, field specific dictionary preprocessing and Huffman coding optimization are combined, a repetition mode of a specific field text is fully utilized, and the high compression rate of 70%-80% is achieved and is far higher than that of a general algorithm by 40%-60%.

This invention discloses a waveform recognition and correction method and system for winding deformation diagnosis, relating to the field of transformer winding monitoring technology. The method includes: synchronously acquiring voltage and current signals and performing preprocessing; identifying waveform disturbance types in the preprocessed signals, classifying disturbance types into harmonic-dominant disturbances, voltage sag disturbances, or composite disturbances; dynamically selecting appropriate waveform processing paths based on the waveform disturbance type identification results; processing the signals using the selected waveform processing paths to correct and reconstruct the signals, obtaining compensated voltage and current waveform data; and constructing a disturbance recognition and path adaptive processing mechanism to solve the processing lag and compensation distortion problems of general algorithms under dynamic signal disturbances, improving recognition accuracy and achieving real-time error correction.

The invention discloses a centralized control station side equipment state early warning method and system based on a general algorithm architecture. The architecture comprises early warning algorithm basic configuration, early warning data acquisition, equipment data effective judgment, equipment early warning algorithm analysis, early warning monitoring display, early warning auxiliary analysis, early warning information cross-regional synchronization and equipment early warning visual analysis. The architecture acquires real-time data, historical data and alarm information of a subscription early warning object according to basic configuration information of various types of equipment early warning algorithms to obtain data required by equipment early warning, filters invalid equipment and measuring point data through equipment data validity judgment, periodically schedules and analyzes various types of early warning algorithms, and performs early warning on the equipment. And pushing early-warning real-time alarms and storing early-warning result data. The architecture is used for realizing active early warning of equipment hidden dangers, and assists equipment monitoring operation and maintenance professionals in finding equipment problems in advance through an early warning monitoring interface, an early warning monitoring assistant, early warning information cross-regional synchronization and early warning visual analysis, and timely disposes and avoids potential safety hazards.

The present application relates to the technical field of optical lens manufacturing, and provides a lens taking method and system based on a general algorithm of lens mold cavity coordinates. The method comprises the following steps: establishing a mechanical coordinate system and taking the mold center fixed on a tool as the origin of the mold coordinate system, obtaining the offset of the mold center in the mechanical coordinate system as the center offset, setting the number of mold cavities N, the mold base ring radius R and the first cavity offset angle θ1, and setting the taking sequence of the cavity number counterclockwise / clockwise, calculating the two-dimensional coordinates under the mold coordinate system for the kth mold cavity, performing a homogeneous translation on the two-dimensional coordinates and the center offset to obtain the target grabbing point under the mechanical coordinate system, grabbing the lens mold cavity at the target grabbing point, and repeating the taking value k from 1 to N to complete the cavity-by-cavity taking of the lenses in the counterclockwise / clockwise taking sequence of the optical lens production mold. The method of the present application can be used for taking lenses from annular multi-cavity molds, and meets the actual needs of high yield, low breakage rate and fast changeover in lens production.

The present invention relates to an adaptive asset management apparatus and method, the apparatus comprising: a memory including at least one instruction; and at least one processor for executing the at least one instruction stored in the memory, wherein the processor identifies an investment time point and an investment cost for each of a plurality of facilities on the basis of time-based management (TBM), identifies a corresponding year in which the investment cost exceeds a preset budget constraint, and changes the investment time point by calculating an optimal solution for the excess amount by which the investment cost exceeds the budget constraint by using a genetic algorithm. Other embodiments are also applicable.

The invention relates to the technical field of optical lens manufacturing and the like, and provides a lens taking method and system based on a lens mold cavity coordinate general algorithm. The method comprises the steps that a mechanical coordinate system is established, the center of a mold fixed to a tool serves as the original point of the mold coordinate system, the offset of the center of the mold in the mechanical coordinate system is obtained and serves as the center offset, the number N of mold cavities, the radius R of a mold core ring and the first cavity deflection angle theta 1 are set, and the reverse / clockwise mirror taking sequence of cavity serial numbers is set; and for the kth mold cavity, calculating a two-dimensional coordinate under a mold coordinate system, performing homogeneous translation on the two-dimensional coordinate and the center offset to obtain a target grabbing point under a mechanical coordinate system, grabbing the lens mold cavity at the target grabbing point, and repeating the value of k from 1 to N to complete the cavity-by-cavity extraction of the lenses in the reverse / clockwise lens extraction sequence of the whole optical lens production mold. The method disclosed by the invention can be universally used for lens taking of the annular multi-cavity mold, and the actual requirements of high yield, low fragment rate and rapid remodeling in lens production are met.

This invention discloses a multi-domain voltage fingerprinting method and system for identifying transformer-household relationships using a synergistic computing module, belonging to the field of power distribution network management technology. The method involves acquiring the voltage time-domain waveform via high-speed sampling using an HPLC synergistic fusion module, followed by adaptive variational mode decoupling, multi-scale permutation entropy-scattering feature joint characterization, and continuous homology topological feature extraction. After confidence-weighted concatenation, transformer-household relationship discrimination is achieved through Riemannian manifold combined with spectral clustering. The system consists of a distributed HPLC synergistic fusion module and a transformer concentrator. The module performs local feature extraction and uploads compressed feature vectors, while the concentrator performs feature fusion, distance calculation, and cluster recognition. This invention introduces continuous homology into transformer-household identification for the first time. Pure voltage-driven operation requires no additional hardware, edge computing significantly reduces communication overhead, and the system boasts high accuracy, strong noise robustness, and adaptability to existing equipment upgrades, improving the level of refined management in transformer substations.

The application discloses a kind of gas sensor array adaptive front fusion detection method, belong to sensor intelligent detection technical field, comprising: data set preparation and pre-processing;Based on the construction of neural network model of deep learning and MASK operator;The training of neural network model based on deep learning and MASK operator;Actual concentration value and array sensor device state are detected using the trained neural network model described above.The application integrates array sensor and interference gas sensor into calculation, the input data is the original data collected from each sensor, and the output data obtained after algorithm fusion analysis is the gas concentration value and the working state of each sensor device, realizing the end-to-end information extraction in true sense;The successful construction of the algorithm helps to promote the construction of gas sensing system general algorithm framework and automated algorithm optimization.

The invention relates to the technical field of power data storage and transmission, and discloses a power utilization data compression method, a power utilization data transmission system, a power utilization data compression device and electronic equipment. Generating a power consumption data matrix based on the original power consumption data; judging whether the data of the power utilization data matrix is valid or not, and generating a matrix mark; effective data are screened based on the matrix marks, the effective data are compressed by adopting a compression algorithm, and compressed data blocks are generated in combination with the matrix marks, domain knowledge is converted into a compression rule, the defect that a general algorithm is insufficient in utilization of structural data features is overcome, and the transmission bandwidth and storage cost of mass power consumption data are remarkably reduced. And meanwhile, the real-time requirement in a high-frequency acquisition scene is met.

The present disclosure relates to a method and apparatus for determining a virtual output of a multi-energy x-ray device. Based on the application for which the x-ray device is being used, a general algorithm can be determined or selected. Inputs received from the x-ray device can be substituted into the general algorithm to generate a virtual output algorithm for the x-ray device. The virtual output algorithm can then be used to calculate a virtual output.

The present disclosure describes a three-way hybrid quantum computing system that includes a classical computer, a quantum annealer, and a gate-based quantum computer. The system combines the unique strengths of each component to more efficiently solve complex computational tasks. The classical computer is responsible for general-purpose tasks, data storage, and control of the quantum components. The quantum annealer is used to solve optimization problems, while the gate-based quantum computer is used to execute more general algorithms. The system converts problems that group the commutants of a Hamiltonian into a quadratic unconstrained binary optimization (QUBO) problem, instructs a hybrid combination of the classical computer and the quantum annealer to solve and validate the results, and then utilizes a hybrid combination of the classical computer and the gate-based quantum computer to deploy the results to determine the results of all terms in each commutant group in the Hamiltonian in parallel. The system provides an output representing a solution to the original problem.

This invention discloses a general algorithm and supporting execution module for stress control and performance improvement of all-solid materials, belonging to the field of stress regulation and structural performance enhancement technology for solid materials. Based on a general adaptive algorithm and a dedicated closed-loop execution module, it achieves full-process control of detection-prediction-optimization-iteration through real-time stress sensing, intelligent decision-making, and global micro-compensation closed-loop regulation, achieving high-precision stress balance convergence for industrial all-solid materials and simultaneously improving structural stability and overall performance. It features a built-in manual intervention channel and hardware safety interlock, providing industrial-grade safety assurance and adaptability to stable operation under extreme conditions. The algorithm is implemented purely in software and can be deployed independently on existing computing platforms, or adapted to external devices or used in conjunction with this module to achieve optimal accuracy. Accuracy levels depend on the performance of external sensors and execution hardware. Overall performance is significantly superior to existing technologies and is suitable for various industrial fields with solid material stress control needs.

In view of the problem that the optimization method based on the subspace is insufficient in precision in the field of full-wave inverse scattering imaging, a kernel subspace optimization method for full-wave inverse scattering high-precision imaging is disclosed, which relates to the fields of machine learning and full-wave inverse scattering. The method combines the robust framework of the traditional subspace optimization method and the adaptive learning advantage of the kernel method, and transfers the iteration of the original subspace optimization method to the reproducing kernel Hilbert space, so that the evolution in the learning iteration process is learned with the least mean square error as the learning criterion. The kernel-based subspace optimization method combines the mature and interpretable framework of the subspace optimization method with the adaptive learning ability of the kernel method, and the method provides a robust and general algorithm for full-wave inverse scattering in the fields of medical imaging, geophysical exploration and nondestructive testing, and can provide accurate reconstruction results.