26 results about "Constrained least squares" patented technology

The invention relates to a red light therapeutic instrument real-time calibration method based on multi-modal data fusion, which comprises the following steps: firstly, collecting data such as optical power, target surface temperature, internal temperature, reflection spectrum and light source-target surface distance in a time window, and unifying time base alignment; performing anomaly elimination, de-noising and normalization on each modal data, and constructing a standardized data matrix; extracting multi-modal features through power spectral density, wavelet packet decomposition, exponential moving average and Kalman filtering, inputting the multi-modal features into a fusion network containing a modal special encoder and a cross-modal attention unit, and generating a red light treatment state vector; the driving current, the pulse width, the duty ratio, the emission angle and the aging compensation factor are solved in real time through constrained least square optimization and are issued to the driving control module, and the red light source is smoothly adjusted in a soft start mode; the system monitors the output power, calibration is completed if the output power meets a threshold value, and otherwise, the next iteration is started. According to the method, power closed-loop correction can be realized, and dose consistency and thermal safety are improved.

The invention discloses a three-dimensional elastic block-fault motion deformation coupling model construction method and system. Comprising the following steps: acquiring geodetic survey data of earth crust movement between earthquakes and an active fault distribution map, and dividing different active blocks; the movable blocks are associated with geodetic measurement data through spatial position attributes, and motion features of different blocks are distinguished; adjacent boundaries are extracted to serve as modeling boundary faults, and discretization processing is carried out; constructing a three-dimensional elastic block-fault motion deformation coupling model including block rotation, internal strain, fault normal velocity discontinuity counteracting, inclined fault positive slippage and negative dislocation response; inverting parameters of the three-dimensional elastic block-fault motion deformation coupling model by adopting a constrained least square method; and according to the coupling coefficient, identifying the potential pregnancy and earthquake concave-convex body on the fault. According to the method, a more complete inter-earthquake three-dimensional elastic model is established, and support is provided for accurate research of active fault coupling characteristics, recognition of gestational earthquake concave-convex bodies and judgment of major earthquake dangerous areas.

The invention provides a gain phase error calibration DOA (direction of arrival) estimation method for a movable antenna array, and aims to optimize local gain and phase error problems caused by calibration errors in an actual system. On the basis of a two-stage structure, firstly, a rotation-invariant signal subspace is constructed, an estimation problem of a gain phase error is converted into a constrained least square problem, a closed-form solution is obtained by using a Lagrangian multiplier method, and joint estimation and compensation of the error are realized. And then, in the second stage, constructing a noise reduction covariance matrix and introducing an estimated gain phase error for compensation, and finishing high-precision direction-of-arrival estimation in combination with a MUSIC algorithm. A simulation experiment shows that compared with an existing method, the self-calibration method provided by the invention shows more excellent direction of arrival estimation performance in a movable antenna array. The method does not need extra error distribution priori knowledge, has good practicability, and is suitable for application scenes needing accurate DOA estimation, such as wireless sensing and radar.

This invention provides a photocatalytic disinfection efficacy evaluation system based on dynamic detection of microbial activity, belonging to the field of photocatalytic material detection technology. It includes a signal acquisition module for real-time acquisition of fluorescence intensity signals of live and dead microorganisms in photocatalytically treated microbial samples; a signal stability control module for calculating the coefficient of variation of the fluorescence intensity signal using a sliding time window and automatically adjusting the injection flow rate of the fluorescent reagent based on an incremental PID control algorithm; a survival rate calculation module for unmixing the stabilized signal using a constrained least squares method based on a pre-calibrated fluorescence signal unmixing matrix to obtain the instantaneous survival rate time series of microorganisms in real time; and an efficacy evaluation module for fitting the time series data to a first-order inactivation kinetic model, solving for the inactivation rate constant, its confidence interval, and goodness-of-fit index. This invention achieves real-time, online, standardized quantitative evaluation of photocatalytic disinfection efficacy.

This invention relates to the field of power grid sensing and data analysis technology, and discloses a two-stage topology identification method, device, equipment, and storage medium for power distribution systems, which improves the accuracy and reliability of power distribution network topology identification. The two-stage topology identification method for power distribution systems includes: acquiring smart meter measurement data and cumulative energy consumption data of inter-meters, and performing preprocessing through current calculation, time alignment, and power conversion; subsequently, constructing a constrained least squares optimization problem to solve for the impedance sub-matrices between smart meters and between smart meters and inter-meters; based on the impedance sub-matrices, using a recursive grouping algorithm to construct a smart meter subgraph, and then successively determining the connection positions of inter-meters to generate the target power distribution network topology.

This invention discloses a method for establishing an automatic berthing model for underactuated ships, relating to automatic berthing and unberthing of ships. The method includes: establishing a three-dimensional geometric model of the ship based on its linesform, propeller linesform, and rudder linesform; establishing a numerical simulation environment for the three-dimensional geometric model using computational fluid dynamics (CFD) software, and conducting numerical simulation experiments on the three-dimensional geometric model within the numerical simulation environment; establishing a mathematical model of the ship's berthing process based on the numerical simulation experiments; and identifying the ship's hydrodynamic parameters using the constrained least squares algorithm based on the mathematical model of the ship's berthing process. This invention, based on ship CFD virtual numerical simulation and constrained least squares algorithm for identifying hydrodynamic coefficients, effectively solves the problems of difficult acquisition of actual ship test data and high test costs. Furthermore, the use of CFD software allows for simultaneous execution of multiple sets and types of tests, effectively improving the efficiency of model establishment.

The invention provides a photocatalytic disinfection efficiency evaluation system based on microbial activity dynamic detection, and relates to the technical field of photocatalytic material detection, the photocatalytic disinfection efficiency evaluation system comprises a signal acquisition module used for acquiring fluorescence intensity signals of live microorganisms and dead microorganisms of a microbial sample subjected to photocatalytic treatment in real time; the signal stability control module is used for calculating a variable coefficient of a fluorescence intensity signal according to a sliding time window and automatically adjusting the injection flow rate of a fluorescent reagent based on an incremental PID control algorithm; the survival rate calculation module is used for carrying out unmixing on the stabilized signals by adopting a constrained least square method based on a pre-calibrated fluorescence signal unmixing matrix so as to obtain an instantaneous survival rate time sequence of the microorganisms in real time; and the efficiency evaluation module is used for fitting the time sequence data to a primary inactivation mechanical model, and solving an inactivation rate constant and a confidence interval and a goodness of fit index thereof. According to the invention, real-time, on-line and standardized quantitative evaluation of the photocatalytic disinfection efficiency is realized.

The invention belongs to but not limited to the technical field of satellite communication, and particularly relates to an OTFS-based fractional Doppler channel estimation method and system in low-orbit satellite communication, and the method comprises the steps: S1, employing a ZC sequence as a pilot frequency, and providing a low-PAPR DD domain pilot frequency structure; s2, analyzing the influence of IDI on a ZC pilot frequency period correlation result, and revealing a module value distribution rule; s3, initial estimation based on correlation peak value calculation is provided, and initial channel parameters can be obtained; and S4, constructing a non-convex constraint least square optimization problem, and performing solution by using an SQP algorithm so as to obtain more accurate channel parameters. According to the DD domain frame structure provided by the invention, the ZC sequence is adopted as the pilot frequency, and the guard interval is set, so that the PAPR can be reduced, and the interference between data and the pilot frequency can be effectively reduced. The module value distribution rule of the correlation result of the pilot frequency period of the ZC sequence under the influence of IDI is disclosed for the first time, and the rule is fully utilized to carry out initial channel estimation, so that the accuracy and reliability of estimation are improved.

The application relates to a converted wave shear wave splitting calculation method and system, equipment and a storage medium, and belongs to the technical field of oil and natural gas geophysical exploration. The method comprises the following steps: respectively pre-processing, sorting and dynamic correction of converted wave radial component data and converted wave tangential component data; synchronously dynamic correcting radial component gather data and tangential component gather data; respectively performing azimuthal division on the obtained radial component data and tangential component data after synchronization; based on the azimuthal division obtained common receiver point radial component and tangential component azimuthal division gather, a phase inversion method and a phase-constrained least square method, the shear wave splitting is calculated according to a vertical layer. The application combines the advantages of the phase inversion method and the least square method, has lower requirements on input data, is suitable for the situation of low signal-to-noise ratio data and irregular acquisition data, and improves the shear wave splitting calculation precision.

The invention relates to the technical field of drilling, in particular to a multi-channel drilling tool system integrated with an in-situ test function, which comprises a multi-channel response collection module used for calling real-time response values of an optical spectrometer, a sound velocity sensor and a resistivity sensor on a drilling tool assembly, and calculating the real-time response values of the optical spectrometer, the sound velocity sensor and the resistivity sensor; and integrating theoretical response values of different fluid components in the preset physical property library to form an observation matrix. According to the method, the multi-dimensional data of the optical spectrometer, the sound velocity sensor and the resistivity sensor are called in parallel, and the non-negatively constrained least square operation is carried out in combination with the preset physical property library, so that the real-time and quantitative analysis of the concentrations of the four components of oil, gas, formation water and filtrate is realized, the accuracy and the anti-interference capability of fluid identification are improved, and the accuracy of fluid identification is improved. Meanwhile, by extracting a time sequence of filtrate concentration and dynamically fitting the time sequence with a radial composite flow function, traditional passive and empirical sampling waiting is converted into active and accurate dynamic prediction, and non-production operation time is shortened.

The invention discloses a soil organic carbon source and stability rapid evaluation method based on a solid 13C nuclear magnetic resonance spectrum and an intelligent analysis system, and the method comprises the following steps: carrying out standard pretreatment on a soil sample, and detecting the soil sample by using a solid 13C nuclear magnetic resonance technology to obtain spectrum data; partitioning the obtained spectrum data according to a standard chemical shift interval, and calculating an integral area proportion of a partition corresponding to each chemical functional group; based on the molecular mixing model, constructing a linear equation set, and solving the mole fractions of the six chemical components in the soil sample by adopting a constrained least square programming algorithm; and calculating the proportions of plant source carbon, microbial source carbon and pyrolytic carbon in the soil sample according to the calculated mole fractions of the six chemical components. The carbon proportion of a plant source and a microorganism source can be effectively distinguished by combining solid 13C nuclear magnetic resonance and a molecular mixed model, and the carbon stability change in the grassland degeneration and recovery process is revealed.

The application discloses a kind of flexible connection constraint under multi-node distributed filtering estimation method, belong to deep space exploration technical field.The implementation method of the application is as follows: by selecting the node with maximum observability in adjacent node, the state estimation thereof is utilized to establish flexible connection constraint, and the constraint least square problem of distributed filtering estimation is formed;The condition that inequality constraint is converted into equality constraint needs to be satisfied is analyzed using KKT condition, the flexible connection constraint is converted from inequality constraint to equality constraint, to avoid optimization process under the influence of inequality constraint to fall into local optimum, by constraint local linearization, the analytical expression of node state estimation mean value and estimation error covariance is obtained, the efficient estimation of flexible lander node state is realized, and the state estimation result conforming to the physical characteristics of flexible connection is generated;In addition, the nonlinear constraint linearization error is reduced by iterative estimation, and the node state estimation accuracy of flexible lander is further improved.

This invention discloses a CNC polishing method for controlling the material removal depth of blade surfaces, comprising: Step 1: obtaining the target removal distribution covering the entire blade surface parameter space; obtaining a mapping relationship model for predicting the tool influence function of the surface; Step 2: within the blade surface parameter domain, establishing a discrete convolution model of the dwell time of discrete points on the polishing trajectory, the target removal distribution, and the predicted tool influence function, introducing upper and lower bounds of the dwell time constraint and a differential penalty term to ensure speed stability, constructing and solving a constrained least squares model for dwell time optimization, thereby obtaining the optimal dwell time, and finally mapping it to the CNC system feed speed according to the trajectory segment length; This invention can customize the removal depth according to the blade surface characteristics and initial error, and significantly improve the controlled accuracy of material removal on complex surfaces while ensuring the stability of CNC machining.

A node positioning method for marine environmental monitoring wireless sensor networks is provided, which jointly estimates node positions and path loss factors by considering the node real-time movement and the path loss and absorption effect of underwater communication; transforms an original non-convex problem into a non-negative constrained least squares framework, and finds the optimal solution of marine node positions by two stages of interior point method and block coordinate update. In the first stage, the problem is re-expressed by using the penalty function according to the interior point method to obtain the approximate solution; and in the second stage, the original problem is transformed into a generalized trust region sub-problem, and the approximate solution obtained by the interior point method is used as the initial estimation, the accurate estimation values of the marine node position and path loss factor are obtained by iterative solutions combined with the block coordinate update.

A mixed signal computing-in-memory (CIM) system solves integer constrained least square (ICLS) optimization problems through parallel analog computation within memory arrays. The system transforms ICLS problems into energy minimization problems solved using Hopfield neural network dynamics. A first CIM component performs iterative optimization using quantized weights stored in memory arrays, with analog current accumulation providing parallel computation and stochastic feedback enabling escape from local minima. A second CIM component evaluates solution quality through distance metric computation. The system converts integer optimization variables to binary representations suitable for neural network processing. Binary neuron states are generated through current comparison and modified with programmable noise injection for global optimization. The system supports constant time operation for predictable latency and dynamic time operation for faster convergence. The CIM approach eliminates data movement bottlenecks while achieving significant speedup and energy efficiency compared to conventional digital processing methods for solving nondeterministic polynomial time-complete ICLS problems.

The present application relates to a kind of ocean ecological environment monitoring wireless sensor network node positioning method, consider the path loss and absorption effect of node real-time movement and underwater communication, the position of node and path loss factor are estimated jointly;Convert original non-convex problem into non-negative constraint least square framework, find the optimal solution of marine node position based on two-stage of interior point method and block coordinate updating method;First stage is according to interior point method using penalty function to restate the problem, approximate solution is obtained;Second stage, the original problem is converted into generalized trust region subproblem, the approximate solution obtained by interior point method is used as the initial estimate of second stage, and the accurate estimation value of marine node position and path loss factor is obtained by combining a block coordinate updating method for iterative solution.The present application solves the problem that positioning error is relatively large due to node real-time movement, unknown path loss factor and absorption effect of underwater signal propagation in the process of underwater acoustic signal propagation, and realizes high-precision and high-robust positioning of marine target node.

The application discloses a construction method and system of a source positioning high-efficiency estimator based on mixed measurement information, and the method steps comprise the following steps: constructing a mixed information non-negative constraint least square framework based on an analysis sensor in a wireless sensor network; obtaining a rough suboptimal estimation based on the mixed information non-negative constraint least square framework; and completing the construction of the high-efficiency estimator based on the rough suboptimal estimation. The original non-convex optimization problem is converted into a non-negative constraint least square framework through linearization calculation. The suboptimal estimation is obtained through the index exchange process between the active set and the passive set by using the linear solution. Meanwhile, in order to reduce the deviation caused by linearization and refine the estimation obtained by linear solution, a deviation reduction method is used, a weighted instrumental variable matrix which is weakly related to noise but strongly related to the data matrix is designed, so that the positioning precision is improved.

The application proposes an intelligent light coordination control system and method based on artificial intelligence, which comprises the following steps: acquiring lamp operation state, environment lighting state and engineering parameters, and generating a basic executable constraint interval; utilizing an adjacency list and coupling weight to cooperatively trim the basic constraint set to generate a system-level cooperative constraint set; within the range limited by the system-level cooperative constraint set, combining the current state and neighborhood trend value, and determining the target control value through constrained least square calculation; and based on the first-order inertia model discretization logic, calculating the actual control value and issuing it for execution. Through dynamic constraint construction and cooperative optimization, the application effectively solves the adjustment conflict problem under multi-lamp coupling, realizes smooth transition and regional uniformity of light control, and improves lighting comfort and energy saving effect.

The application discloses a method, device and medium for improving remote sensing estimation of ground object abundance in a pixel. The application belongs to the technical field of remote sensing imaging detection, and solves the technical problem that the full constraint least square method is affected by a sensor spatial point spread function, and errors exist in obtained ground object abundance. The method comprises the following steps: S1, collecting remote sensing image data; S2, estimating ground object abundance in a remote sensing image pixel by using an improved full constraint least square method; in the step S2, the improved full constraint least square method introduces the influence of the sensor point spread function when estimating the ground object abundance in the remote sensing image pixel, and improves the mixed pixel value; and S3, obtaining the estimation result of the ground object abundance in the remote sensing image pixel. The method is suitable for the technical field of remote sensing imaging.

The embodiment of the invention relates to a sparse constraint least square seismic time-frequency analysis method and device, and the method comprises the steps: carrying out the complex processing of a real seismic channel, and constructing a complex seismic channel; constructing a diagonal matrix with a data weight as a window function as a diagonal line, and initializing a model weight into a unit matrix; constructing an Euler formula kernel matrix formed by different cut-off frequencies and time ranges in a time domain; applying the data weight constraint to the complex seismic trace, and applying the data weight constraint and the model weight constraint to the Euler formula kernel matrix; and constructing an error function among a constrained Euler formula kernel matrix, a model parameter and a constrained complex seismic trace, applying L1 regularization sparse constraint to the model parameter, and performing inversion by applying a greedy iterative shrinkage threshold method to obtain a sparse constrained least square seismic time-frequency analysis result. Therefore, the seismic signal can be efficiently and accurately decomposed to obtain the high-precision time-frequency spectrum.

A wireless positioning method based on biased RSS measurement, which uses biased RSS measurements collected by all anchor nodes to construct a high-dimensional space convex hull, so that the unknown parameter vector containing the target point position and sensor bias falls into the convex hull as much as possible, and the vertices of the convex hull are used as virtual points. Secondly, a constrained least squares estimation problem is established based on the obtained virtual points, and the convex combination coefficients are solved. Finally, the estimation of the unknown parameter vector is represented by the convex combination of the corresponding virtual points, and the target position estimation and sensor bias estimation are obtained. The positioning accuracy of this method under the condition of biased RSS measurement is significantly better than that of existing positioning methods that do not consider bias.

The application discloses a kind of based on big data analysis's electric pole intelligent maintenance prediction method, comprising the following steps: obtaining the monitoring data corresponding to target electric pole, constructs multi-source state feature sequence;Residual state feature sequence is constructed;Carry out credible constraint and purification treatment, form current state distribution sample set;Carry out working condition mapping compensation and inverse offset correction, form counterfactual health reference distribution sample set;Relative unconstrained least square importance fitting algorithm RuLSIF is executed, and relative density ratio offset result is calculated;The continuous monitoring period of target electric pole is divided into phase before disturbance, phase in disturbance and phase after disturbance recovery, respectively calculate the relative density ratio offset result corresponding to each phase, obtain migration offset state;Determine whether irreversible hidden danger migration of target electric pole occurs, output maintenance prediction result.The application improves the accuracy of electric pole abnormal state recognition and the foresight of maintenance prediction.

The invention discloses a method and equipment for improving abundance of ground objects in a remote sensing estimation pixel and a medium. Belongs to the technical field of remote sensing imaging detection. The technical problem that errors exist in acquired surface feature abundance due to the fact that a fully constrained least square method is affected by a sensor space point spread function is solved. The method specifically comprises the following steps: S1, collecting remote sensing image data; s2, estimating the abundance of the ground objects in the pixels of the remote sensing image by adopting an improved fully-constrained least square method, introducing the influence of a sensor point spread function when the abundance of the ground objects in the pixels of the remote sensing image is estimated by adopting the improved fully-constrained least square method, and improving a mixed pixel value; and S3, obtaining a surface feature abundance estimation result in the remote sensing image pixel. The method is suitable for the technical field related to remote sensing imaging.

The application discloses a kind of decoupling control methods of regional dimming, comprising: analyzing input image to obtain the target brightness of each dimming region, to establish target brightness matrix;The target brightness matrix is input into backlight decoupling neural network model to obtain the best dimming brightness matrix;And the best dimming brightness is used to control backlight module to emit light.Backlight decoupling neural network model The training data set used in the modeling process is associated with constrained least squares problem, and the target is to find the best dimming brightness matrix from multiple dimming brightness matrices, so that the difference between the diffusion result and the target brightness matrix is minimized.Diffusion result is obtained by substituting dimming brightness matrix into diffusion formula, and diffusion formula is established according to the point spread function of each light emitting element of backlight module in each dimming region.

This invention discloses a quantitative measurement method and system for the three-dimensional pattern of impervious surfaces, comprising the following steps: reconstructing spatially continuous Landsat imagery; extracting the feature set with the highest accuracy in measuring impervious surfaces as the optimal feature set; estimating the abundance information of impervious surfaces using fully constrained least squares; gridding the city using a three-dimensional spatial grid; constructing quantitative indicators using building vector data and DEM data, and determining the weight of each indicator in the measurement of the three-dimensional pattern of impervious surfaces using the entropy method, calculating the development score of the three-dimensional pattern of impervious surfaces, and describing the three-dimensional pattern of impervious surfaces. This invention quantitatively describes the three-dimensional pattern of impervious surfaces, which helps to understand the three-dimensional spatial distribution of cities and provides a reference for urbanization patterns.

The invention relates to a hyperspectral-thermal imaging diabetic foot multi-point risk assessment method and system based on condition number adaptive unmixing, and belongs to the technical field of medical images and artificial intelligence. According to the system, a plantar hyperspectral data cube and a thermal image are synchronously collected, a linear model of hemoglobin chromophores and a scattering substrate is established through the expansion correction Beer-Lambert law, non-negative constraint least square with robust loss is adopted for solving, and a physiological parameter graph and a pixel-level confidence graph are output; performing temperature calibration, foot segmentation and double-foot registration on the thermal image, combining Getis-Ord Gi * space statistics and DBSCAN anomaly cluster detection, and calculating a thermal anomaly index; physiological and thermal features are extracted from a plurality of ROIs such as heels and metatarsal bones based on a foot anatomy template, and an interpretable machine learning model is input to output risk grades, early warning positions and confidence coefficients. The method is oriented to noninvasive and automatic diabetic foot early risk early warning, and has higher robustness and repeatability in a complex nerve-ischemic lesion scene.