34 results about "Linear algorithm" patented technology

The invention discloses an augmented reality head-up display system and a display method thereof, relates to the technical field of auxiliary driving, and aims at complex road conditions and diversified driving requirements. The comprehensive fusion analysis of the driving environment and the driver state is realized by sequentially executing the four steps of environment perception and multi-source data acquisition, situation analysis and information complexity judgment, dynamic content screening and display strategy generation and real-time monitoring and self-adaptive feedback closed loop. The method comprises the following steps of: marking and normalizing multi-dimensional sensing data to form an environment factor and a driving load factor, and judging the current information complexity by using a nonlinear algorithm; thirdly, screening the most critical content from the information pool according to the priority, and performing refined arrangement according to the display position and style; and if the HUD configuration deviation is monitored and detected, dynamic strategy updating is executed, so that the AR-HUD always ensures the accuracy and the conciseness of information display in different scenes, and the driving safety and the user experience are improved.

The invention discloses a drainage pipe network operation intelligent monitoring system based on multi-source data fusion, and relates to the technical field of drainage pipe network intelligent monitoring. According to the method, through collaborative design of the dynamic weight matrix and the ChebNet-LSTM fusion model, limitation of a traditional algorithm on nonlinear feature capture and time-space correlation analysis of the drainage pipe network is broken through; the dynamic weight matrix is combined with node topological data and abnormal case data, so that the association strength between the nodes can be dynamically adjusted, and the complex working condition of the pipe network can be accurately adapted; according to the ChebNet-LSTM model, space correlation features are extracted through a ChebNet layer, long-term and short-term time sequence dependence is captured through an LSTM layer, the problem that a traditional linear algorithm cannot fit a non-linear scene is solved, the limitation of single-dimension feature analysis is broken through, pipe network data space-time collaborative modeling is achieved, multi-task output can be achieved while the calculated amount is reduced to meet the real-time monitoring requirement, and the real-time monitoring efficiency is improved. And the abnormity pre-judgment accuracy and the tracing precision are greatly improved.

The invention relates to a Stewart six-dimensional force sensor, six branch chain rods distributed in an inclined manner are arranged between an upper platform and a lower platform, each branch chain rod comprises an upper rigid rod, an upper cross flexible hinge, a measuring unit, a lower cross flexible hinge and a lower rigid rod which are sequentially connected from top to bottom, the upper rigid rod is fixedly connected with the upper platform, and the lower rigid rod is fixedly connected with the lower platform. And the lower rigid rod is fixedly connected with the lower platform. A calibration device is used for calibrating a Stewart six-dimensional force sensor. The invention relates to a Stuart six-dimensional force sensor calibration method, which comprises the steps of off-line calibration and training and on-line measurement and calculation. The Stewart six-dimensional force sensor has the advantages of being large in rigidity, high in bearing capacity, stable in structure, fast in dynamic response, high in precision, good in isotropy and the like, and solves the problem that a traditional linear algorithm cannot process complex nonlinear errors. The measurement precision is reduced due to serious inter-dimensional coupling; and the complexity of a high-precision model is too high and the high-precision model is difficult to run in real time in an embedded system.

The present disclosure relates to a method and apparatus for determining a signal detection network. The method includes: determining sample communication parameters for communication between a sample transmitter and a sample receiver via a sample intelligent reflector surface (IRS) (S101); and training an initial neural network based on a training sample set composed of the sample communication parameters to obtain a signal detection network, wherein the input of the initial neural network is the sample communication parameters, and the output of the initial neural network is an estimate of the transmitted signal of the sample transmitter (S102). According to the present disclosure, the initial neural network can be trained based on the training sample set to obtain a signal detection network. Since the neural network is not a linear algorithm and is not limited to operating on linear relationships, it can be effectively used for signal detection even in complex communication environments with many nonlinear factors, and a high-precision estimate of the transmitted signal can be obtained.

The invention relates to the technical field of automatic control and intelligent driving, and discloses a precise macro-micro composite driving system control method based on a self-learning nonlinear PID algorithm. The method comprises the steps of collecting real-time operation data to judge nonlinear dynamic characteristics, obtaining a deviation value, extracting nonlinear characteristics and adjusting control parameters, switching to a self-learning mode to update a model, compensating interference and the like. By monitoring and compensating the nonlinear characteristics and external interference of the system in real time, the response speed and control precision of the system are improved, the dependence on external experience parameters is reduced, and the system adapts to complex and changeable working environment requirements.

The invention discloses a spectrum simulation nonlinear algorithm and system based on DMD modulation. By establishing a working state, wavelength distribution, an energy weight and a target spectrum matrix, constructing an operation rule and considering DMD surface illumination nonlinearity, automatic fitting of different skewness spectrum shapes is realized, skewness errors are reduced, and spectrum simulation accuracy is improved; a matrix operation system reduces fitting complexity and improves simulation efficiency. And a BP neural network optimized by a genetic algorithm is adopted to fit a spectrum minimum unit, so that the self-adaption and optimization capability of the algorithm is enhanced, and the spectrum is accurately fitted. A 450nm-1000nm spectrum shape fitting experiment is completed by using a specific experimental device, the effectiveness and the reliability are verified, and the method has wide application prospects in multiple fields and can meet spectrum simulation requirements of different scenes.

The application discloses a quantum linear solving method and device based on quantum adiabatic linear algorithm, and a medium, and relates to the technical field of quantum computing. The method comprises the following steps: obtaining a linear system to be processed, and constructing a Krylov subspace matched with the linear system; determining a subspace equation set of the linear system based on the Krylov subspace according to the linear system and the Krylov subspace; constructing a quantum circuit corresponding to the quantum adiabatic linear algorithm according to the subspace equation set, and solving an approximate solution of the linear system in the Krylov subspace according to the quantum circuit. The application can combine the subspace method with the quantum adiabatic linear algorithm, perform dimension reduction processing on the original high-dimensional linear equation set in the quantum adiabatic linear algorithm through the subspace method, so that the dimension of the linear equation set actually solved in the iteration process is smaller than the dimension of the original linear equation set. Therefore, the demand of the quantum discrete adiabatic linear algorithm for computing resources is reduced, and the quantum discrete adiabatic linear algorithm can be implemented in an NISQ chip and a classical computer.

The application discloses a feed formula determination method and device, electronic equipment and storage medium. According to a plurality of feed information and animal growth required nutrient structure information, a target constraint condition is determined, and a preset first linear algorithm and a second linear algorithm are used to generate first feed formula information and second feed formula information that satisfy the target constraint condition according to the feed information, so as to preliminarily obtain the feed formula through the linear algorithm. And the preset genetic algorithm is used to combine and iterate the first feed formula information and the second feed formula information until the preset iteration termination condition, and third feed formula information is obtained, so that the genetic algorithm is used to effectively avoid the situation that the linear algorithm falls into a local optimal solution and cannot obtain a global optimal solution, the nonlinear space solution of linear programming processing is made up, and the reliability of the feed formula optimization result is improved.

An apparatus, system and method for providing a may include at least a phacoemulsification surgical console having a customizable non-linear custom aspiration mode. The console may include at least an aspirator; a foot pedal; and a non-transitory computing code resident on a computing memory associated with a computing processor which, when executed by the processor, causes to be executed the steps of: receiving a percentage actuation of the foot pedal; calculating, including from a non-linear algorithm, a percentage actuation for the aspirator corresponded to the received percentage foot pedal actuation; and dictating the calculated percentage aspirator actuation to the aspirator.

The application relates to the field of power electronics, and particularly provides a LLC resonant converter soft-start control method and device and an LLC resonant converter, wherein the method comprises sequentially performing three-stage soft-start control, wherein in the first stage, a fixed dead zone is used to gently establish resonance for a first time length; in the second stage, a non-linear algorithm based on multiple hiccup periods and a decreasing quantity increasing with periods is used to smoothly adjust the dead zone; and in the third stage, the dead zone is decreased to a target value at a two-stage speed of slow first and then fast; fault detection including at least start timeout and input voltage abnormality is performed in parallel in the whole process; start completion is judged based on multiple conditions of total time, dead zone value and voltage ratio; and when the driving dead zone is updated, protection of not being lower than a safety minimum value is forced to be performed. Through the above common control, the converter is smoothly and quickly started, and the start impact and voltage stress are effectively inhibited.

The invention discloses an observable method and system for a unified log index and tracking data, and the method comprises the steps: collecting various types of original observation data, carrying out the unified formatting, constructing a column type storage structure, and carrying out the hierarchical storage; then optimizing data layout by utilizing column-level metadata and a query mode feature library, and implementing a multi-level index strategy; meanwhile, through data cleaning, structured enhancement and contextual information association, a causal atlas between heterogeneous observation data is established, and observation data with rich contexts is generated; and finally, a high-performance query processing system is constructed based on a unified query engine and a sub-linear algorithm, and progressive result return and visual analysis are realized. According to the method, the data retrieval efficiency and the analysis depth are remarkably improved, operation and maintenance personnel can more quickly position the root cause of the system problem, and the average time of fault diagnosis is effectively shortened.

The invention relates to the technical field of data analysis, in particular to an Internet of Things time sequence analysis method based on a self-adaptive hypergraph neural network, which comprises the following steps: acquiring a plurality of human activity identification data, determining a selected data volume according to data fluctuation, determining activity label similarity according to information association degree, and obtaining a time sequence analysis result; according to the activity label similarity, human activity identification data is selected as training data according to the overall similarity and the selection similarity or according to the amplitude change rate; according to the motion complexity of the training data, determining whether to increase and adjust the scale feature level to determine a target scale feature level; and the features of the points are connected through Tensor, and classification is carried out through Linear to obtain a prediction category result corresponding to the training data. The method can improve the recognition precision of the human activity recognition task.

The embodiment of the invention discloses a fitting precision detection method and device and a storage medium. The fitting precision detection method comprises the steps that at least two images are acquired and preprocessed, and the at least two images are images of different angle areas of the plane object; for each image, capturing according to a preset capturing frame, performing linear fitting on the captured area according to a Hough linear algorithm to obtain a fitted linear segment, screening the fitted linear segment according to a preset screening condition, and taking the screened linear segment as the protective film edge of the corner area corresponding to the current image; and calculating the laminating precision of the protective film according to the position information of all the screened straight line segments.

The invention relates to the technical field of image processing, in particular to a compensation dispersion transfer entropy-based causal analysis method, and aims to accurately capture an instantaneous causal relationship and improve the accuracy of a constructed brain network. The method comprises the following steps: acquiring an original time sequence, and normalizing the original time sequence by adopting a normal cumulative distribution function to obtain a normalized time sequence; and performing dispersion processing on the normalized time sequence, and endowing each normalized data point with an integer in a preset integer range through a linear algorithm to obtain a dispersion time sequence. And performing information transmission quantitative analysis on the dispersive time sequence based on the information entropy, and evaluating the uncertainty degree of information flow transmission from the source system to the target system and the uncertainty degree of information flow transmission from the source system to the target system under the constraint of a preset condition. And introducing a compensation mechanism to optimize the uncertainty evaluation result.

The invention discloses a vacuum infusion cyclic extrusion pressure control device and method based on a double-layer linear algorithm, and belongs to the technical field of composite material manufacturing, the device comprises double sets of independent hardware systems, a double-pipeline control system and a pressure control box; the method comprises a pressure control method and a compaction rate control method. The compaction rate control method comprises the steps that in the cyclic extrusion process, the compaction rate is accurately controlled by adjusting the opening proportion of a vacuum electromagnetic valve of a second pipeline of the control device, and a double-layer linear algorithm is adopted; the pressure control method adopts segmented pressure control, comprises high-pressure section control and low-pressure section control, and is controlled by a first pipeline or two pipelines of the control device together. The pressure control precision reaches + / -0.1 kPa, and the compaction rate control precision is within + / -6%; the control precision can be kept within the atmospheric pressure change range of 97.5-102.5 kPa, and the system adapts to different geographical and meteorological conditions; the full-range control from 0.5 kPa to atmospheric pressure and the compaction rate control from 1 kPa / s to 5.5 kPa / s can be realized.

The invention provides a simulation engine multi-parameter cooperative control method, device and equipment, a medium and a program product, and the method comprises the steps: obtaining the load change information of a simulation engine; under the condition that the load change information is larger than a preset threshold value, the operation state of the simulation engine is determined to be a transient state, a closing instruction is sent to a control module of an exhaust gas recirculation valve, and a first control instruction is sent to a control module of a variable nozzle turbine; the opening degree of the variable nozzle turbine is reduced to a first preset opening degree and is increased to a second preset opening degree after being kept for a preset time length, and the opening degree of the variable nozzle turbine is used for determining transient air inlet data transmitted to the simulation engine; and a second control instruction is sent to the fuel injection module, so that transient fuel injection data of the fuel injection module is controlled to reach first preset fuel injection data within the total control duration by using a stage-linear algorithm, and the fuel-oxygen equivalence ratio, obtained through simulation based on the transient air inlet data and the transient fuel injection data in the transient state, of the simulated engine is reduced.

The invention discloses a multi-region sheet resistance optimization calculation method and system based on voltage scanning, and belongs to the field of electronic device design and control. By inputting area parameters and power density parameters of a multi-region device and combining voltage scanning and sheet resistance constraint verification, global optimization matching of sheet resistance of each region is realized. Compared with a traditional local optimization method, global solution space search is achieved based on the O (N) linear algorithm, local optimal traps are avoided, the calculation efficiency is improved by more than 10 times, the square resistance uniformity standard deviation can be reduced to below 0.3 and is verified in distributed thin film heater design, and the effect of remarkably improving temperature uniformity and device reliability is achieved.

The invention discloses a method for predicting the minimum miscible pressure of CO2 and crude oil by using machine learning, and the method comprises the steps: carrying out the high-precision modeling based on a linear SVM algorithm, building a correlation prediction formula, and constructing a correlation prediction model; and training and optimizing the correlation prediction model by using the collected data, and establishing a stacked integrated model in combination with a random forest regression model RF and the correlation prediction model to realize accurate prediction of the CO2-crude oil minimum miscible pressure MMP. According to the method, the time and the cost for determining the CO2-crude oil MMP value can be reduced to the maximum extent. Therefore, a reliable CO2-crude oil MMP value can be provided through prediction, and the miscible phase type of CO2 oil displacement is determined, so that the design of a CO2 oil displacement scheme is improved, the optimal design of an injection-production scheme is guided, and the crude oil recovery rate is increased.

Computer processing and outcome prediction systems and methods used to generate algorithm time prediction polynomials, inverse algorithm time prediction polynomials, determine race conditions, determine when a non-linear algorithm can be treated as if it were linear, as well as automatically generate parallel and quantum solutions from classical software or from the relationship between monotonic attribute values.

This invention provides a drive circuit for inverter loads, relating to the field of inverter technology. It includes a control module, a drive module, a power supply module, a signal acquisition and processing module, and a load matching module. The load matching module has impedance matching, tuning matching, and shaping filtering functions, providing universally adaptable drive capabilities for most types of motors and ultrasonic transducers. The power supply module uses high-speed MOSFETs, and the drive module uses a high-speed drive chip to adapt to the high-speed MOSFETs, offering a wide frequency range. A DDS waveform generation module is used to achieve PWM wave output at different frequencies, with extremely low frequency change steps and smooth transitions. Combining a variable step-size algorithm and a LADRC algorithm, it can automatically and accurately lock the operating resonant frequency of the inverter load within a certain frequency range. The employed linear LADRC algorithm adjusts the behavior of the drive circuit through real-time measurement and estimation, adapting to the variations and uncertainties in the complex working environments of motors and ultrasonic transducers.

The invention discloses a rapid approximate block coding quantum circuit construction method and a linear equation set solving method and device. The solving method comprises the following steps: obtaining a linear equation set Ax = b; constructing a time-dependent Hamiltonian H (f (s)) of adiabatic evolution, wherein the eigenstate of the initial Hamiltonian H (0) is expressed as bgt; the eigenstate of the target Hamiltonian H (1) is expressed as A-1bgt; ; a discrete adiabatic evolution quantum circuit is constructed, the discrete adiabatic evolution quantum circuit has an evolution operator UH (f (s)) corresponding to the time-dependent Hamiltonian H (f (s)), the evolution operator UH (f (s)) continuously acts on the discrete adiabatic evolution quantum circuit, and the initial quantum state corresponds to the eigenstate bgt of the adiabatic evolution initial Hamiltonian H (0); the final quantum state corresponds to the eigenstate A-1bgt of the adiabatic evolution target Hamiltonian H (1); . The invention provides a specific quantum circuit implementation method of sparse Hamiltonian fast approximate block coding, and realizes an adiabatic quantum calculation linear algorithm based on fast approximate block coding to complete the solution of a linear equation set.

The application provides a blood flow signal denoising method based on PSO-EMD optimized wavelet threshold, which comprises the following steps: collecting photon signals scattered by tissues by using a photoelectric detector, processing the signals by a digital correlator, and obtaining blood flow signals by N-order linear algorithm calculation; decomposing the blood flow signals by EMD to obtain IMF signals; selecting a wavelet threshold denoising algorithm, and dynamically screening the threshold by using PSO, taking the signal-to-noise ratio as a target fitness function, and selecting optimal parameters by considering the target adaptation principle and the effective convergence principle of the PSO algorithm; removing noise from high-frequency IMF components in the IMF signals according to the wavelet threshold denoising algorithm of the optimal parameters, and superimposing and reconstructing the high-frequency IMF components, low-frequency IMF components and residual components to obtain denoised blood flow signals. The application can effectively remove noise and maximize the details of the effective information of the blood flow signals.

This invention relates to the field of wireless communication transmission technology and discloses a data transmission method for weak network environments. The method includes: acquiring the channel impulse response of the first and second symbol periods of a wireless channel; extracting multipath delay parameters, signal amplitude parameters, and discrete time axis coordinates; mapping the parameters to a spatiotemporal parametric coordinate system; generating first and second channel feature surfaces; calculating the curvature variance of the first channel feature surface; when the curvature variance is lower than a preset correlation distortion threshold; extending the second channel feature surface along the evolution gradient trend to generate a predicted channel feature surface; thereby determining the equalizer tap coefficients and compensating for data symbol distortion. This invention enhances transmission reliability in highly dynamic weak network environments by locking the gradient divergence path generated by the linear algorithm in a Doppler shift environment through geometric surface physical continuity constraints.

The invention provides a rolling bearing diagnosis method and system based on an orthogonal projection mechanism and time domain orthogonal constraint. The method comprises the following steps: acquiring a vibration signal of a rolling bearing through a sensor; an orthogonal dictionary or an over-complete dictionary is adaptively selected according to the vibration signal intensity and frequency distribution characteristics, and dictionary parameter optimization and dictionary construction are completed according to a preset rule; based on the selected dictionary type, utilizing an ODMP or TOMP algorithm to calculate a sparse coefficient; reconstructing a signal based on the sparse coefficient; and analyzing an envelope spectrum of the reconstructed signal to extract a fault characteristic frequency and a harmonic wave thereof, modulating a sideband, and completing rolling bearing fault type determination. According to the method, pseudo-inverse calculation and repeated projection operation of a traditional OMP algorithm are eliminated through an orthogonal projection mechanism and time domain orthogonal constraint, the calculation complexity of the ODMP algorithm is reduced to be linear, the complexity of the TOMP algorithm is remarkably reduced, the accuracy of sparse coefficient estimation is ensured, and the operation time is greatly shortened; the method is suitable for rolling bearing early weak fault detection.

The application provides a solving method and system for phase difference wave front detection, and the method comprises the following steps: collecting a plurality of target images with phase difference through an imaging system; obtaining an estimated value of an imaging wave front phase through the target images and a neural network; and obtaining the imaging wave front phase by using a nonlinear algorithm based on the estimated value. The method provided by the application can quickly and accurately solve the optical wave front phase, has the advantages of high solving precision, fast processing speed, good method robustness and strong practicability, can directly give a better estimated value of the imaging wave front phase without iteration operation which consumes a large amount of time, and the trained network also maintains high precision for real images through simulation image training of the neural network, avoids the harsh requirement of collecting a large amount of real images as training data, and has good application feasibility.

This invention discloses a blockchain-based information security detection system, relating to the field of information security technology. It includes: an off-chain monitoring node that uses a non-linear algorithm to calculate the dynamic reputation score of the monitored entity in real time. A trigger signal is generated only when the reputation score crosses a preset reporting threshold. This trigger signal then backtracks to the local cache and extracts the Merkle root hash of key events and related data, assembling them into a standardized evidence package and constructing a notarization transaction. The notarization smart contract responds to this transaction, recording the evidence package in the blockchain state ledger and reading the current block height as the unique temporal trust benchmark for the evidence package for associated storage. This invention, through the combination of off-chain dynamic risk threshold triggering and on-chain consensus time anchoring, achieves an automated, high-fidelity conversion from dynamic risk states to tamper-proof on-chain evidence, and establishes a network-wide consensus time trust root that does not rely on a single node's clock.

The invention discloses a harmonic power flow analysis method considering background harmonic voltage, which comprises the following steps of: establishing a harmonic equivalent circuit comprising a background harmonic voltage source and a plurality of nonlinear loads, and establishing a node admittance matrix according to the equivalent circuit; a method of establishing a harmonic equivalent conductance and harmonic equivalent susceptance model is used to decompose harmonic current at a node, and harmonic current reflecting nonlinear characteristics of a harmonic source is extracted. The nonlinear load harmonic current is used as node injection harmonic current, each harmonic voltage at the node is solved by adopting a current injection type retaining nonlinear algorithm, and when the convergence condition is met, the harmonic voltage value of each node under the condition of considering the background harmonic voltage can be solved. And analyzing the transmission characteristics of harmonic waves in the power distribution network system according to the obtained node harmonic voltage. According to the method, the harmonic load flow calculation can be performed on the power distribution network under the condition of considering the background harmonic voltage to obtain an accurate harmonic voltage value in the system, so that the harmonic transfer characteristic is analyzed, and the application is wide.

The invention solves the problem of the difficulty of acquiring subsea variables by using a widely available surface variable, namely the pressure upstream of the choke valve. As surface variables have an unfavorable dynamic for use in conventional anti-slug controllers, which are based on the linear PID algorithm, the proposed controller uses a hybrid fuzzy-PID architecture, which compensates for the unfavorable dynamic of the controlled variable by means of heuristic interventions in the control action generated by the PID part of the controller. The heuristic action of the proposed controller allows it to be more robust than a conventional controller, even with a relatively slow control action, which makes it possible to apply the proposed algorithm in systems whose choke valve activation is slow. Thus, the two largest impediments for installation of conventional controllers, lack of subsea measurement and choke valve slowness, do not pose a problem for application of the proposed algorithm. This allows the proposed anti-slug controller to be used in offshore production systems without the need for physical intervention in the subsea or surface facilities.

The present invention discloses a method for controlling the constant current accuracy of a resistor-capacitor circuit, which relates to the technical field of resistor-capacitor circuits. Specifically, it includes: Step S1, constructing a resistor-capacitor circuit topology and establishing a small-signal model based on the resistor-capacitor circuit topology; Step S2, adopting a double closed-loop control of an outer current loop and an inner voltage loop. In the outer current loop, a reference voltage signal is generated by a PI controller, and in the inner voltage loop, current tracking is achieved based on a non-linear algorithm; Step S3, introducing a dynamic compensation mechanism to suppress accuracy errors through feed-forward compensation and parameter adaptation. This application adopts a double closed-loop control of an outer current and an inner voltage, and dynamically compensates the current error through feed-forward compensation and parameter adaptation. Through accurate calculation of the duty cycle and closed-loop correction, the constant current accuracy control of a multi-branch resistor-capacitor circuit is realized.

The invention discloses a data prediction modeling method based on a large language model, electronic equipment and a medium, and the method comprises the steps: obtaining first and second prediction data with labels, and setting a screening amount N; screening out N first tags related to the second tag through a large language model; respectively extracting first and second knowledge features corresponding to the first and second prediction data; calculating the correlation between the first knowledge features to construct a first priori knowledge matrix, and calculating the correlation between the second knowledge features and the first knowledge features to construct a second priori knowledge matrix; and constructing a regression prediction model based on a Bayesian linear algorithm, taking the first priori knowledge matrix and the second priori knowledge matrix as priori knowledge, taking the first prediction data as the input of the regression prediction model, and taking the second prediction data as the output of the regression prediction model, thereby realizing data prediction from the first prediction data to the second prediction data.