28 results about "Sum squared error" patented technology

The invention discloses a power distribution network traveling wave fault positioning method and a related device, and relates to the technical field of power distribution networks, and the method comprises the steps: collecting a transient signal after a fault at a preset key monitoring point of a power distribution network, extracting an initial traveling wave head from the transient signal, and measuring the moment when the initial traveling wave head reaches other preset monitoring points in the power distribution network, obtaining traveling wave arrival time between the key monitoring point and each monitoring point; the time difference between the traveling wave arrival time is used as the actually measured traveling wave arrival time difference to construct a fault positioning optimization model with the purpose of minimizing the error sum of squares of the theoretical traveling wave arrival time difference and the actually measured traveling wave arrival time difference; and solving the fault positioning optimization model through a pre-constructed chaos particle swarm optimization algorithm to obtain a traveling wave fault positioning result of the power distribution network. The method solves the problems that the positioning effect is not ideal due to the fact that the prior art is prone to falling into local optimum and is poor in adaptability to a complex topological structure, and the convergence speed cannot meet the requirement for rapid power supply recovery.

The invention belongs to the technical field of unmanned aerial vehicle aircraft design, and particularly relates to a longitudinal aerodynamic parameter identification calculation method in a closed-loop state and a large angle-of-attack range, and the method comprises the steps: building a flight simulation model; performing aerodynamic characteristic simulation based on a flight simulation model to obtain an aerodynamic load value, comparing the aerodynamic load value with data tested in an actual flight experiment, and calculating and minimizing an error sum of squares; dividing the whole angle-of-attack range into a plurality of sub-intervals, respectively obtaining the optimal aerodynamic parameters of each sub-interval based on the minimum sum of squares of errors, inputting each optimal aerodynamic parameter into the flight simulation model, and carrying out model accuracy verification; and drawing an aerodynamic characteristic curve according to each optimal aerodynamic parameter, and verifying the applicability of the flight simulation model. The method has universality and is suitable for longitudinal aerodynamic coefficient parameter identification of different types of unmanned aerial vehicles, the parameter identification efficiency is improved, and meanwhile the applicability of the model under the complex flight working condition is enhanced.

The application discloses a lightweight indoor 3D laser ground extraction algorithm method, system and medium. The method comprises the following steps: acquiring indoor 3D laser radar data; obtaining coordinate values based on a laser radar coordinate system according to the indoor 3D laser radar data; sending the coordinate values to a preset plane equation to obtain a ground plane normal vector; sending the ground plane normal vector to a first preset coordinate system to establish a constraint to obtain a comprehensive error value; and accumulating squares of values in the comprehensive error value to obtain a comprehensive error sum of squares; extracting a minimum value in the comprehensive error sum of squares, and setting a coordinate point corresponding to the minimum value as an optimal current pose. The application is suitable for various lasers, and by giving a preset threshold, whether it is 16 lines, 32 lines or 128 lines, the algorithm accuracy and robustness are ensured, and the accuracy of recognized ground points is improved.

The application provides a ship type classification prediction method and system based on K-means and XG-Boost, acquires ship data and carries out pretreatment, then adopts a K-means clustering algorithm to respectively cluster each kind of data in the pretreated ship data to obtain multiple clusters, calculates error sum of squares of all data in each cluster, calculates the classification number of each kind of data according to the error sum of squares, selects a certain classification number by using an elbow method, marks the ship type of all clustered ships according to the classification number, then takes the ships with marked ship types as training set samples, adopts an XG-Boost classification algorithm to train the training set samples to obtain multiple classification prediction models, verifies the multiple classification prediction models to obtain an optimal classification prediction model, and predicts the ship types of all ships in the world according to the optimal classification prediction model and marks the ship types. The application can accurately classify all ships in the world, and can avoid overfitting and underfitting of the model while ensuring the accuracy.

This invention discloses a data visualization monitoring method and system for heating networks in heating station supply areas, specifically relating to the management field. The method includes synchronously collecting multi-dimensional data, establishing a heating network spatial coordinate system to locate the data collection nodes, and storing the coordinate data in a data storage center. It employs a K-means clustering algorithm to extract three typical heating patterns and simultaneously evaluates the matching degree of user heating patterns. An optimization cycle is set to extract optimized samples, and the gradient descent method is used to minimize the sum of squared errors between model predictions and actual values, updating relevant model parameters. Based on data analysis results, this invention generates three targeted decision-making suggestions: resource allocation, inspection priority, and heating guidance. Combined with historical data tracing and operational pattern mining, it provides comprehensive support for short-term operational optimization and long-term planning of the heating network, effectively improving the efficiency of heating resource utilization, reducing operational risks, and minimizing resource waste, thus exhibiting significant economic and management benefits.

The invention belongs to the technical field of unmanned aerial vehicle design, and particularly relates to an open-loop state large-angle-of-attack range longitudinal aerodynamic coefficient identification calculation method, which comprises the following steps of: comparing simulation data with test data in an actual flight test, calculating an error sum of squares of the simulation data and the test data, and minimizing the error sum of squares; dividing the whole attack angle range into a plurality of sub-intervals, respectively obtaining the optimal aerodynamic parameters of each sub-interval based on the minimum sum of squares of errors, inputting each optimal aerodynamic parameter into the flight simulation model, comparing the optimal aerodynamic parameters with test data for verification, and executing the next step when the difference value between the optimal aerodynamic parameters and the test data is within a set range; and drawing an aerodynamic characteristic curve according to each optimal aerodynamic parameter, verifying the applicability of the flight simulation model, and optimizing the robustness of the flight simulation model. The method is suitable for longitudinal aerodynamic coefficient parameter identification of different types of unmanned aerial vehicles, the parameter identification efficiency is improved, and meanwhile the applicability of the model under the complex flight working condition is enhanced.

The invention relates to a tunnel unmanned aerial vehicle cluster mobile ad hoc network and navigation method and system in a closed space. Determining the number of unmanned aerial vehicles most suitable for covering the current ground user group and the target position of each unmanned aerial vehicle by adopting a clustering algorithm based on error sum of squares and contour coefficient value optimization so as to realize an unmanned aerial vehicle cluster mobile ad hoc network; and a dynamic adaptive potential field algorithm based on multi-agent reinforcement learning is adopted to plan an air route for each unmanned aerial vehicle to arrive at the corresponding target position. According to the method, the coverage range and the service quality of the unmanned aerial vehicle system are effectively improved, and efficient path planning is realized in a dynamic and complex environment.

The application relates to a privacy protection recommendation method and system based on matrix decomposition and belongs to the technical field of privacy protection recommendation. The method comprises the following steps: obtaining a historical rating matrix of a user; performing matrix decomposition on the historical rating matrix to obtain a shared rating matrix; sending the shared rating matrix and a scored item set to a recommendation server to generate public parameters, a decomposition matrix and a shared value of the decomposition matrix; generating inner product triplets and multiplication triplets by using a triplet server; calculating errors between real ratings and predicted ratings by using the recommendation server to generate an error shared matrix; optimizing the error shared matrix until the error squares of all the error shared matrices are less than a threshold value; calculating partial derivatives of a loss function on the decomposition matrix to obtain a gradient shared matrix; updating the decomposition matrix to obtain a user decomposition shared matrix; and generating a privacy protection recommendation result of the user by using the user decomposition shared matrix. The application provides strong protection in the aspect of protecting user privacy.

The invention relates to the technical field of power systems and renewable energy scheduling, and discloses a small hydropower station cluster planning and grouping method and device and a medium, and the method comprises the steps: obtaining multi-source data of each small hydropower station, and constructing a small hydropower station cluster feature space model; constructing a mathematical effectiveness index and a field rationality index, and establishing a k-means clustering algorithm multi-dimensional evaluation index system; executing an improved k-means + + clustering algorithm, and generating clustering schemes and evaluation results under different K values; constructing a derivative evaluation index based on an error sum of squares, distributing an index weight, calculating a comprehensive score corresponding to each K value, and determining an optimal clustering k value; and outputting a small hydropower station cluster grouping scheme corresponding to the optimal clustering number. According to the method, adaptive grouping of small hydropower station clusters is realized through a clustering algorithm driven by a small amount of data, and a small hydropower station cluster division scheme with reasonable spatial distribution, unified basin attributes and coordinated operation characteristics is provided for regional power grid dispatching.

A deformable block registration improvement method based on historical information correction comprises the steps that an image is layered according to scenery information, and key areas are marked according to application requirements and a layering result; random block selection and deformation calculation in the key area: randomly selecting blocks in the key area, adjusting the size of the selected blocks, and calculating an initial deformation matrix of the selected blocks by adopting a normalized error sum of squares and an affine model; correcting a deformation result based on a historical deformation field: constructing historical deformation field data, extracting historical deformation data of the current block, and correcting an initial deformation matrix through a dynamic weight model; and performing new deformation field fitting: applying consistency constraint to the correction block, fitting a global deformable field by adopting thin-plate spline interpolation, outputting after filtering optimization, and updating a historical deformation field database. According to the method, the problems of low registration precision and error accumulation of deformable blocks in a multi-moving-target scene are solved through marking key areas by image layering, performing targeted random block selection, performing historical deformation field correction and performing consistent deformation field fitting.

The invention discloses a power system state estimation Hessian matrix method based on optimal control, and belongs to the technical field of power system state estimation. According to the method, a node power equation for state estimation of the alternating-current power distribution network is established, and a Jacobian matrix and a Hessian matrix of node power unbalance to node voltage and phase angle are deduced; a state estimation optimization model with the minimum sum of squares of node power measurement errors as a target is constructed, the power system state estimation Hessian matrix method based on optimal control is provided, the state estimation convergence is improved, meanwhile, the calculation speed is guaranteed, and compared with a traditional Newton-Raphson state estimation method, the method has the advantages that the calculation speed is improved, and the calculation efficiency is improved. The state estimation method is few in iteration times and good in convergence.

The present invention discloses a railway subgrade settlement prediction method based on the IGM-PSVR model. First, the subgrade settlement data is collected and preprocessed. Then, the data is subjected to a level ratio check, a corresponding differential equation is established, and the time response function is solved. The background value of the gray model is further optimized, and an improved gray model IGM(1,1) is constructed to obtain the prediction results of the model. Then, the particle swarm optimization algorithm is used to optimize the hyperparameters of the support vector regression (SVR) model, and a particle swarm optimized support vector machine model is constructed to obtain the prediction results of the model. Finally, the inverse sum of squared errors method is used to assign different weights to the two-step prediction results, and finally the combined prediction results are obtained. Experiments show that the combined model has higher prediction accuracy than the single model, can better reflect the trend of subgrade settlement changes, and plays an important role in ensuring the safety of railway transportation.

This invention relates to an automatic focusing method for laser galvanometers based on a polynomial regression algorithm, including variable definition, data collection of sample data within different power ranges from 10% to 80%, and data preprocessing; employing a piecewise polynomial regression model, and adding an L2 regularization term to the model to prevent overfitting; and minimizing the observed value D. i The model parameters are estimated using the sum of squared errors between the predicted and actual values. The model is used to predict values ​​in real time. When the absolute value of the deviation ∈ between the predicted and actual roundness values ​​D exceeds the threshold Th, the galvanometer height H or laser power P is dynamically adjusted to make D approach the target value D. target =0; the system is based on the optimal height H * The Z-axis position of the galvanometer is adjusted in real time to focus the laser beam at the optimal position; the deviation between the actual value D and the predicted value ∈ is periodically collected to update the regression model coefficients. This invention enables precise focusing of the laser beam, improving product consistency and yield.

The invention relates to a straight line and circular arc combined convex shape optimization method for a cylindrical roller, and aims to solve the problems that in existing logarithmic shape modification fitting, circular arc segmentation connection points depend on experience setting, and adaptability is poor. According to the method, a corresponding line-arc combination curve model is constructed based on a logarithmic modification curve, and the line-arc combination curve is iteratively optimized through an optimization method to obtain an optimal line-arc combination curve, so that the optimal line-arc combination curve approaches the logarithmic modification curve. The modeling process comprises segmentation point setting, connection point generation, arc parameter solving and curve construction, the optimization target is to minimize fitting error quadratic sum, and the geometric rationality and convexity of the curve are constrained and guaranteed. The adopted optimization algorithm is a sequence least square quadratic programming (SLSQP) algorithm. According to the method, fitting precision and engineering applicability are remarkably improved, and the method has good universality and expansibility, is suitable for modification design of various roller bearings and has engineering application value.

The application discloses a kind of closed loop identification method, device, terminal equipment and storage medium based on nuclear power steam generator model, by adopting autoregressive model structure, according to several historical operation data, the initial system model of nuclear power steam generator is structured, then least square method and auxiliary variable method are used, to determine the observation variable in initial system model, to be identified parameter, residual error and the parameter initial value of to be identified parameter, therefore, the present application can overcome the nonlinear problem when the nuclear power steam generator is modeled at present, and the problem of low quality of closed loop operation sampling data of nuclear power steam generator.Finally, by minimizing the way of prediction error square sum, the iteration model parameter is updated using prediction error method, effectively improve parameter identification precision, finally obtain the target value of to be identified parameter, to build the system coupling model that can be used to carry out real-time monitoring and control to the operating state of steam generator.

The application provides a robot joint friction model parameter identification method, device and equipment. The method provided by the application comprises the following steps: a speed-related friction model, a temperature-related friction model and a load-related friction model are established to constitute a robot joint friction model; a plurality of groups of friction characteristic data of a robot joint in different working states are acquired, and initial value ranges of various parameters in the model are set based on expert knowledge and prior physical knowledge to constitute an initial parameter search space; an error square sum of model prediction is taken as a fitness function, the initial parameter search space is optimized based on the plurality of groups of friction characteristic data by using a genetic algorithm to obtain a candidate parameter search space; and an optimization function with the optimization objectives of minimizing a prediction error and minimizing a parameter physical deviation is constructed for the robot joint friction model.

The invention relates to a polynomial regression algorithm-based automatic focus searching method for a laser galvanometer, which comprises the following steps of: defining variables, acquiring sample data in different power sections within 10-80%, and preprocessing the data; a piecewise polynomial regression model is adopted, and an L2 regularization item is added into the model in order to prevent the model from being over-fitted; estimating model parameters by minimizing the sum of squares of errors between the observed value Di and the predicted value; when the absolute value of the deviation epsilon between the predicted value and the actual roundness value D exceeds a threshold value Th, dynamically adjusting the galvanometer height H or the laser power P to enable D to approach a target value Dtarget = 0; the system adjusts the Z-axis position of the galvanometer in real time according to the optimal height H *, so that the laser beam is focused at the optimal position; and periodically collecting a deviation epsilon between an actual value D and a predicted value, and updating a regression model coefficient. According to the invention, the laser beam can be accurately focused, and the consistency and the yield of products are improved.

The invention belongs to the technical field of unmanned aerial vehicle design, and particularly relates to a longitudinal aerodynamic parameter identification calculation method under a closed-loop state and a small attack angle range, and the method comprises the steps: simulating the flight state change process of an unmanned aerial vehicle under the condition of the closed-loop state and the small attack angle range based on a flight simulation model; dynamic characteristics of the aircraft under different longitudinal pneumatic conditions are obtained; performing parameter identification calculation on the dynamic characteristics of the aircraft under different longitudinal pneumatic conditions to obtain flight simulation data; iterative optimization is performed on the longitudinal aerodynamic model parameter combination set through a genetic algorithm, and an optimal longitudinal aerodynamic model parameter combination is obtained by combining the sum of squares of errors between flight simulation data and actual flight test data; and simulation verification and error analysis are carried out on the optimal longitudinal aerodynamic model parameters, and verification is completed when a preset precision requirement is met. The influence of the control law on the flight test in the flight process is considered, and the identification of the longitudinal aerodynamic coefficient parameters in the corresponding closed-loop state is more accurate.

The application discloses a wind power output clustering method and system based on a DTW-Kmedoids algorithm, relates to the technical field of wind power output scene generation, and comprises the following steps: interpolating, reducing the dimension, normalizing and smoothing processing of wind power historical data; adopting a dynamic time warping (DTW) algorithm to perform distance measurement, adopting a Kmedoids clustering algorithm to perform iteration, obtaining a clustering result, and outputting the center curve of each cluster; calculating the error sum of squares of each cluster for the clustering results of different cluster numbers, and adopting an elbow rule to determine the optimal cluster number; generating a typical wind power output scene curve according to the clustering result under the optimal cluster number, and establishing a clustering effect evaluation index system. The application introduces the dynamic time warping distance algorithm, can adapt to the nonlinear change of the wind power output curve, improves the accuracy of similarity measurement through dynamic matching of time series, makes the generated typical scene more representative, makes the clustering center point more representative, and further makes the stability of the generated wind power output scene better.

The invention discloses a multi-peak Gaussian fitting concave cavity blade tip gap measurement method and system, and belongs to the technical field of concave cavity blade tip gap measurement, and the multi-peak Gaussian fitting concave cavity blade tip gap measurement method comprises the steps: S1, obtaining an analog voltage signal reflecting the size of a blade gap; s2, establishing a double-Gaussian model of the concave cavity blade; s3, establishing an error sum of squares model; and S4, parameters in the error sum of squares model are adjusted by using an iterative algorithm, and an iteration termination condition is that the error sum of squares is smaller than a preset threshold, and the obtained parameter value is the parameter of the double-Gaussian model. By the adoption of the technical scheme, the double-Gaussian model and the error quadratic sum model are utilized, and the blade tip clearance of the concave cavity blade can be measured rapidly and accurately.

The invention relates to a high-speed circular sawing machine sawing method and system based on a least square method, and the method comprises the following steps: S1, collecting the non-uniform point cloud data of the edge of a workpiece through a laser scanner, and generating a workpiece edge discrete sampling point set containing N three-dimensional coordinate points; s2, performing segmented cubic spline interpolation processing on the discrete sampling point set to generate a continuous and derivable cutting reference curve; s3, a mathematical model with the sum of squares of the notch track errors as a target function is constructed, the saw blade rotating speed n, the feeding speed v and the sawing depth d serve as optimization variables of the model, and sawing force F is introduced to serve as a constraint condition; s4, on the basis of the mathematical model and the cutting reference curve, the saw blade rotating speed n, the feeding speed v and the sawing depth d are adjusted through an iterative algorithm, the sum of squares of cut track errors is converged under the condition that the constraint condition of sawing force F is met, and an optimal technological parameter combination (n, v and d *) is obtained; and S5, controlling the saw cutting blade to work according to the optimal process parameter combination.

The present invention relates to the technical field of micro-thrust measurement of aerospace micro-thrusters, and discloses a micro-thrust dynamic identification method and system. The identification method comprises: constructing a dynamic equation of a thrust measurement system, rewriting it into a state space form and integrating it to obtain a theoretical system response model at any time; rewriting the theoretical system response model into a numerical model in a step-by-step integral recursive format; constructing an actual system response sequence and combining it with the dynamic equation of the thrust measurement system to construct an empirical model, and rewriting the empirical model; configuring a micro-thrust increment expression, applying the expression to rewrite the numerical model and the empirical model, and further rewriting them into an augmented form; predicting the predicted output at future moments in the time domain; constructing and solving a cost function to minimize the sum of squares of the error between the predicted output and the expected output in the actual system response sequence, while taking into account the smoothness of the micro-thrust input, so that the predicted micro-thrust corresponding to the predicted output is equal to the measured micro-thrust corresponding to the expected output.

The invention relates to the technical field of digital assembly of parts, and discloses a large-scale assembly fixture measurement reference thermal deformation compensation method and device, and the method comprises the steps: obtaining a real-time temperature when a laser tracker measures a coordinate value of a common observation point through a temperature sensor, constructing a temperature field, and carrying out the measurement of the coordinate value through a temperature sensor; then calculating the thermal deformation displacement of each common observation point through a finite element method, compensating the theoretical coordinate point set, registering the compensated and corrected theoretical coordinate point set with the measurement coordinate point set, solving a least square function about the sum of squares of distance errors by adopting a singular value decomposition method, and calculating the thermal deformation displacement of each common observation point; the rotation matrix and the translation vector when the point set in the measurement coordinate system is transferred to the unified coordinate system are obtained, and the station transfer coordinate value is obtained, so that the station transfer precision of the laser tracker is improved, and the method has the outstanding advantage of carrying out rapid compensation aiming at the temperature influence.

The invention discloses an urban rail transit station classification method and system based on multi-feature fusion. The method comprises the following steps: constructing a network topological relation; constructing a passenger flow characteristic index system of the station, and calculating passenger flow characteristic parameters; calculating static characteristic parameters of the site, and standardizing the static characteristic parameters; according to the passenger flow characteristic parameters and the static characteristic parameters, carrying out fusion and dimension reduction processing on the station characteristic parameters to obtain a fused station characteristic data set; determining an initial clustering center according to the fused site feature data set, and carrying out site clustering based on a k-means algorithm; calculating the sum of squares of errors of clustering categories corresponding to different k values and centers of the clustering categories, and determining an optimal clustering number; and according to a cluster division result corresponding to the optimal clustering number, outputting categories corresponding to all stations. According to the method, the time-varying passenger flow characteristics and the static topology characteristics of the rail transit stations are comprehensively considered, the stations are classified through fusion of the two types of characteristic parameters, and accurate classification of the stations is achieved.

The invention discloses a sound source localization method based on reference particle swarm optimization in the technical field of sound source localization, and aims to solve the problems that the traditional particle swarm optimization is easy to fall into local optimum, global exploration and local development are difficult to balance, the convergence speed is low and the like. Acquiring an actual value of time difference of arrival data of the signal received by each sensor; taking an error sum of squares between an actual value and a theoretical value of the time difference of arrival data as a main fitness function; and executing a reference particle swarm optimization algorithm, and performing global optimization on the sound source position. Wherein the reference particle swarm optimization algorithm constructs a dynamic reference particle set, designs a nonlinear adaptive weight mechanism, automatically adjusts the balance of global and local search based on the number of iterations and population dispersion, and introduces a sub-fitness function evaluation module at the same time. According to the method provided by the invention, millimeter-level positioning precision can be realized, and the convergence speed is improved by 80% compared with that of a traditional particle swarm optimization algorithm.

This invention discloses a method for analyzing the machining accuracy of CNC machine tools based on discrete sampling and transient response. It establishes a time decay model for the CNC machine tool, collects input and output data, and uses a hybrid particle swarm optimization (PSO) and gradient descent algorithm to identify model parameters. Machining errors are discretely sampled using a sampling period T. The error sample values ​​at each sampling time are used as input and convolved with the impulse response model. The convolution output is divided into steady-state and transient outputs, and a recursive equation incorporating historical transient summations is established. With the minimum sum of squared machining errors as the optimization objective, a PSO algorithm is used to search for the sampling period and tool feed gain that minimize the objective, and the CNC machine tool control parameters are set accordingly. This invention establishes the response model, the transient relationship between control input and output without changing the hardware. By optimizing control parameters, surface ripple is suppressed, and machining errors are reduced. The model is identified solely based on input and output data.

The application discloses a kind of dynamic characteristic characterization performance acquisition methods of pulse pressure sensor, belongs to pressure sensor metrological calibration field.The application provides a kind of dynamic characteristic characterization performance acquisition methods of pulse pressure sensor, utilizes high-order constant coefficient differential equation with zero output to characterize the dynamic response characteristics of pulse pressure sensor, with the minimum residual error square sum as the basis to carry out system identification, the quantitative differential equation obtained can reflect the influence of measurement system drift;Compared with high-order constant coefficient differential equation of the same order without zero output, smaller system error and random error can be obtained.

The invention discloses an intelligent learning behavior analysis and intervention system and a reminding method, and belongs to the technical field of electric digital processing. The system comprises a data acquisition module, a hybrid analysis module and an intervention execution module. The data acquisition module collects user learning behavior data, such as a preview completion condition, a video watching behavior, an in-hall test answer record and the like. The hybrid analysis module firstly adopts an improved K-Means clustering algorithm to carry out grouping on users, an objective function is that the users are divided into different categories by optimizing the objective function, and the formula is used for measuring the clustering effect. The smaller the SSE (sum of squares of error) is, the closer the clustered sample points to the respective clustering center is, and the better the clustering effect is. By minimizing SSE, sample points in each cluster can be gathered together as closely as possible, so that learning modes and behavior characteristics of different users can be identified more accurately.