54 results about "Dependent parameter" patented technology

This document relates to training of machine learning models such as neural networks. One example method involves providing a machine learning model having one or more layers and associated parameters and performing a pretraining stage on the parameters of the machine learning model to obtain pretrained parameters. The example method also involves performing a tuning stage on the machine learning model by using labeled training samples to tune the pretrained parameters. The tuning stage can include performing noise adjustment of the labeled training examples to obtain noise-adjusted training samples. The tuning stage can also include adjusting the pretrained parameters based at least on the labeled training examples and the noise-adjusted training examples to obtain adapted parameters. The example method can also include outputting a tuned machine learning model having the adapted parameters.

The invention relates to a dynamic modeling method for an attitude control system under a dynamic fault. The method comprises the following steps: determining a modeling hypothesis of a nonlinear dynamic model; establishing a coordinate system used for modeling; rocket configuration parameters and engine related parameters are obtained; building displacement and acceleration descriptions of the three types of structures on the rocket; establishing a rigid body translation equation, a rigid body rotation equation, a rocket body elastic vibration equation, a propellant shaking equation and a sensitive element measurement equation; summarizing the five established equations to form a nonlinear dynamic equation set; and performing small-disturbance linearization expansion on the nonlinear dynamic equation set on the basis of the nominal reconstructed trajectory to obtain a linearized small-deviation equation set suitable for controller design and stability analysis. According to the method, the influence of the thrust drop fault is comprehensively considered, and the model can be used for carrying out relatively accurate dynamic response analysis and reconstruction scheme design under the fault.

The invention provides a multivariate linear fitting deep well crustal stress inversion method and system with multi-working-condition characteristics, and the method comprises the steps: carrying out the simulation modeling of a deep well region through modeling software FLAC3D, endowing a model with related parameters, and obtaining the general situation of the model of a mine region; respectively applying normal stresses in x, y and z directions and shear stresses on x y, y z and z x surfaces to the mine area model in different forms by using FLAC3D; and stress solving is carried out on different working conditions. A multiple linear fitting regression method is adopted for the obtained stress solving result in MATLAB, and a linear fitting regression equation is obtained; and carrying out FLAC3D inversion on the data to obtain a required whole model crustal stress field. According to the invention, the technical problem of low accuracy of inversion operation is solved.

The invention discloses a development method and system of a software development platform, and relates to the technical field of software developing.When the system runs, change records in a code library are extracted, related parameters in code changes are analyzed in detail, a code change set delta C is formed, a dependency relationship module further constructs and updates a dependency relationship graph G in real time, and the dependency relationship graph G is generated; the influence analysis module calculates the influence range of code change by using a dependency graph G, evaluates the influence of the change on a reference object and a subclass, generates a change influence index Idirect, matches the change influence index Idirect with a preset code influence evaluation threshold D, and formulates a code change influence evaluation strategy scheme; and the feedback response module updates data and the dependency graph G according to the scheme content, and notifies a developer of related change conditions, so that the risk of human errors and missed change is reduced, the potential influence caused by code change can be evaluated and responded in time, and the stability and consistency of a code library are ensured.

The invention provides an ABS parameter analysis method, system and device based on a variable step size track and a medium, and relates to the technical field of vehicle control, and the method comprises the steps: obtaining a reference performance index value through a simulation test, and employing a single factor adjustment method to screen out preliminary screening parameters which have effective influences on the performance; then, the parameter intervals are equally divided into multiple sections, and the performance change rate of each interval is calculated to set the weight, so that the importance of different intervals is reflected; and then, normalization processing is combined with interval weights to perform variable-step-size Morris trajectory simulation sampling, so that the number of sampling points is effectively reduced, and the analysis efficiency is improved. And finally, through cross-working-condition integration processing, global maximum sensitive value and sensitivity fluctuation calculation, global key parameters, working condition dependent parameters and stable sensitive parameters are identified, a solid foundation is provided for subsequent accurate parameter calibration, the development period is shortened, the cost is reduced, and the adaptability to different working conditions and the flexibility of parameter optimization are enhanced.

To provide a feature capable of improving the estimation accuracy of a magic formula model used as a tire characteristic model.SOLUTION: This tire characteristic model identification method comprises: a pre-processing step; a first identification step; and a second identification step. In the pre-processing step, a first parameter group is identified in a basic formula of a magic formula model, and a second parameter group is derived using a detailed formula of the magic formula model on the basis of the first parameter group. In the first identification step, setting the second parameter group derived in the pre-processing step as an initial value and a representative parameter in the second parameter group variable, identification is carried out using the basic formula and the detailed formula. In the second identification step, setting the second parameter group updated in the first identification as an initial value and at least one of a load dependent parameter, a camber angle dependent parameter, and an inner pressure parameter variable, identification is carried out using the basic formula and the detailed formula.SELECTED DRAWING: Figure 3

A computer system that computes parameters associated with voxels in a sample is described. During operation, the computer system may obtain information specifying the MR measurements. Then, the computer system may determine an a priori regularizer using a pretrained neural network. For example, the a priori regularizer may correspond to a population of individuals. In some embodiments, the a priori regularizer may correspond to an average person in the population. Moreover, the computer system may compute the parameters based at least in part on the MR measurements, a model of sample physics and the a priori regularizer, where computing the parameters includes solving an inverse problem for the parameters based at least in part on the MR measurements.

The invention describes a method of determining a wear contribution (Ncyc) of at least one bearing of a wind turbine (41), the method comprising: performing a computational simulation of a mechanical behavior of at least part of the wind turbine (41) comprising the bearing for each of a plurality of given operating conditions (28, 29) of the wind turbine across a predetermined simulation time window (8); determining, on the basis of the simulation, at least one mechanical torque (Mx, My), in particular two mechanical torques, which act on the bearing, across the simulation time window (8); determining a time sequence (22) of at least one force-dependent variable (21) across the simulation time window (8) on the basis of the mechanical moments (Mx, My) across the simulation time window (8); determining a time sequence (24a,...) of stress-dependent variables (23) across the simulation time window (8) on the basis of the force-dependent variable time sequence (22) and in particular on the basis of the pitch position (14); the stress-dependent variable time series (24a,...) is evaluated in order to obtain a wear contribution (Ncyc) dependent on the respective operating conditions (28, 29) and the simulation duration.

The application relates to the technical field of system reliability analysis, in particular to a mixed Copula system reliability analysis method based on failure modes, which comprises the following steps: acquiring multi-failure mode degradation data, combining Bayesian inference to estimate parameters of a basic Copula function, rotating and transforming the basic Copula function to construct a mixed static Copula model, judging and optimizing the mixed static Copula model based on root mean square errors, extracting a time-varying correlation parameter sequence by using a sliding window, smoothing and denoising the time-varying correlation parameter sequence by using Kalman filtering, and constructing a mixed dynamic Copula model, and calculating the reliability of the system at different operation time points in combination with a failure mode limit state function. The application can represent the non-symmetrical tail correlation between the multiple failure modes and the time evolution law thereof, and improve the accuracy, stability and engineering applicability of system reliability analysis.

The application discloses a kind of intelligent park infrastructure intelligent monitoring system, it is related to the technical field of intelligent monitoring, including fusion module, prediction module and optimization module, set state related parameter, characteristic data is obtained by fusion, state prediction model is obtained by training, state prediction model is optimized.The application fusion module fuses the characteristic value of multiple state related parameters, reduces the complexity and delay of data processing, improves the monitoring efficiency through automated data processing and model optimization, the characteristic value of multiple related parameters is fused by fusion module, avoid the limitation of single parameter, optimize state prediction model, can dynamically adjust model parameters according to actual operation data, further improve the prediction accuracy of model, can early warning potential failure risk, take measures in time to maintain and repair, reduce the probability of failure occurrence, improve the reliability and stability of power supply facilities.

The application discloses a zero-pole matching discretization application method of a wave trap based on Taylor approximation, which comprises the following steps: determining the type and parameters of the wave trap, obtaining a wave trap transfer function and a transfer function parameter calculation formula according to the principle of the wave trap; discretizing the transfer function of the wave trap by using a zero-pole matching method to obtain an expression of the discrete transfer function of the wave trap; according to the expression of the discrete transfer function of the wave trap, related parameters are approximated by using a Taylor formula to obtain a simplified discrete transfer function, and complex operation is converted into basic operation; based on the implementation mode of the wave trap in a micro control unit, a difference equation is obtained through z inverse transformation, the discretized wave trap based on Taylor approximation is written into the micro control unit, and the effect of the wave trap is verified. The application realizes the complex transfer function of the wave trap, simplifies the calculation of the discrete transfer function, reduces the dependence on a mathematical function library, facilitates the setting of model parameters such as the wave trap, and improves the performance of a control system.

The invention discloses a Qos regulation and control method, device and equipment for a log input size threshold value, and the method comprises the steps: obtaining related parameters of different time points influencing the log input size threshold value, and inputting the related parameters as input vectors into a Transform; performing global time sensing and coding on the input vector by using an encoder to generate a context-related global feature vector, and performing dimensionality reduction and compression on the global feature vector by using a decoder to obtain an optimized feature vector; utilizing a first decision-making machine to decide and output the adjustment direction of the threshold value according to the optimized feature vector, and utilizing a second decision-making machine to decide and output the adjustment amplitude of the threshold value according to the optimized feature vector; and adjusting the current log input size threshold according to the output adjustment direction and the adjustment amplitude. Therefore, the problems that an existing dynamic regulation and control scheme of the log input size threshold is limited by the complexity of a manual writing algorithm, and personal use habits of a user cannot be considered are solved.

The invention provides a high-speed bearing fault diagnosis method for an adaptive parameter completely generalized Gini coefficient. The method comprises the following steps: 1, establishing different state data sets of a bearing; 2, decomposing the signal to obtain a modal component; 3, selecting a part of modal components to calculate a complete generalized Gini coefficient, namely performing sparse processing on the signal; 4, establishing a self-adaptive optimization model, taking the minimum Pearson correlation as a target function, calculating the correlation of the feature fault information in the step 3, and carrying out self-adaptive selection on FGGI correlation parameters a and p; 5, bearing fault feature data are obtained according to the optimal parameters obtained in the step 4; 6, processing the fault feature data, and removing redundant information; 7, establishing a deep extreme learning machine classification model; 8, performing classification and identification on the deep extreme learning machine; according to the method, the FGGI related parameters are adaptively selected, so that the accuracy of fault feature information after subsequent processing can be ensured.

Approaches for top speed dependent operational parameter configuration are disclosed. An operational parameter to be used by an autonomous vehicle is received. Operational modes and corresponding operational regions are determined based on the received operational parameter. A first set of autonomous vehicle parameters are determined and set based on the received operational parameter. A second set of autonomous vehicle parameters are determined and set based on the first set of autonomous vehicle parameters.

This invention provides a method and system for fitting device model parameters, relating to the field of semiconductor device modeling and parameter optimization technology. The method includes: acquiring initial information of a target device model, the initial information including parameters to be fitted; inputting the initial information into the target model to obtain fitting guidance parameters for the parameters to be fitted, the fitting guidance parameters being relevant parameters corresponding to each fitting stage, and the target model being a pre-trained large language model; after the fitting guidance parameters pass verification, fitting the parameters to be fitted at each fitting stage according to the fitting guidance parameters, obtaining the initial fitting result for each fitting stage; dynamically correcting the fitting guidance parameters according to the initial fitting result for each fitting stage to obtain corrected fitting guidance parameters; and fitting the parameters to be fitted at each fitting stage according to the corrected fitting guidance parameters to obtain the target fitting parameters for the parameters to be fitted.

The invention discloses a variable universe fuzzy PID control method for a vacuum suction system, and belongs to the field of industrial control, and the method comprises the steps: building a theoretical model of the vacuum suction system, carrying out the simulation based on the theoretical model, constructing an initial PID controller, determining the initial parameters of the initial PID controller, and setting the related parameters of fuzzy processing; according to the vacuum degree deviation at the current moment and the vacuum degree deviation change rate, calculating to obtain a discourse domain telescopic factor, adjusting a basic discourse domain of the signal by using the telescopic factor, performing fuzzy processing based on a fuzzy rule table to obtain a deviation value of a PID parameter, and adjusting an initial parameter of an initial PID controller by using the deviation value of the PID parameter to obtain an initial PID controller; and finally, the system uses the adjusted PID controller to adjust the duty ratio of the vacuum pump so as to control the vacuum degree of the vacuum suction system. According to the invention, PID parameters in the vacuum suction system can be dynamically adjusted, and stable control of the vacuum suction system in a complex environment can be completed with fewer parameters.

Embodiments of the present disclosure provide a shared library-oriented fuzzer automatic generation method and device, which comprises: extracting an API function set requiring to be subjected to fuzz testing from source code and a header file of a target shared library; identifying control flow dependence and data flow dependence of an API interface based on the API function set, and constructing an API dependence graph; judging parameter types and deducing dependence relationships between parameters according to the API dependence graph, and distinguishing the parameters into independent variable parameters and state-dependent parameters; generating an executable fuzzer through combination scheduling based on function-to-function calling and dependence relationships in the API dependence graph, and embedding a LibFuzzer semantic coding structure into the fuzzer; and continuously tracking a coverage path during execution of the fuzzer, and preferentially selecting an unexecuted path for testing.

The present application provides a premium calculation model test method, system, device and storage medium, the method comprises the following steps: obtaining preset model test parameters, preset model dependent parameters and theoretical premium value; according to the preset model configuration parameter, the effective model test parameter is extracted from the preset model test parameter, the effective model dependent parameter is extracted from the preset model dependent parameter, and the input parameter combination is obtained; the input parameter combination is input into the preset test program, and the test result returned by the test interface is received; the actual premium value corresponding to the policy test data is extracted from the test result, and the test score of the target premium calculation model is obtained according to the actual premium value and the theoretical premium value, combined with the preset error model score contrast relationship. The present application can simultaneously utilize data samples in different scenes to realize double verification of the target premium model, and does not need manual help, greatly reduces the investment of test labor, improves the test efficiency and test quality.

The invention relates to a method for screening key parameters in the production process of a hot-rolled coil plate, and belongs to the technical field of steel rolling optimization control methods. According to the technical scheme, key parameters are collected and sorted, and target parameters are analyzed and determined according to quality requirements; relevant parameters in the rolling process of the hot-rolled coil plate are collected to construct a data set; carrying out correlation analysis on all parameters in the data set and the selected target parameter, and obtaining a correlation coefficient; key parameters in the production process of the hot-rolled coil plate are screened, and feature engineering processing is conducted on a data set obtained after data arrangement; performing feature conversion on parameters belonging to category type data in the data set, and converting feature values into numeric type data; and performing parameter cross validation on the data set. The method has the beneficial effects that the parameters influencing the product quality can be quickly selected and modified, the difficulty of adjusting the production parameters is greatly reduced, the parameter adjusting speed is increased, and the product yield is remarkably increased.

The present invention discloses an optimization method for dimensional parameters and machining parameters, with the optimization object being design parameters; constructing the objective function of the integrated optimization model for shaft bridge design and manufacturing and the constraint function of design parameters; optimizing the shaft bridge dimensional parameters and shaft bridge machining parameters through sequential quadratic programming according to Formula 1, i.e., the integrated optimization model for shaft bridge design and manufacturing: solving to obtain the optimal shaft bridge dimensional parameters and shaft bridge machining parameters. Through iterative calculation according to Formula 1, the present invention can calculate the optimized values of each design parameter from the relevant parameters of the initial existing shaft bridge, which not only meets the constraint threshold but also makes it the smallest, so that the finally obtained shaft bridge dimensional parameters and shaft bridge machining parameters can achieve the optimal cost on the premise of ensuring the performance target. The present invention also provides a shaft bridge manufactured using this optimization method.

Various aspects are directed to improving and simplifying the design of modes like CCLM or LIC based on a parametric model of correlation of neighboring samples. The proposed modifications relate to the way in which the parameters of said parametric model are derived and the way in which prediction tools based on said parametric model, contained in a codec, are designed in a unified and simplified manner. In one embodiment, a method proposes a simplification of the cross-component linear model procedure for deriving linear parameters. It is proposed to derive the parameters as parameters of a straight line passing through two points corresponding to the minimum and maximum luminance values ​​among all luminance neighboring reconstructed samples, instead of the least mean square method.

This invention discloses a method for applying parameter perturbations during missile-aircraft separation, comprising: S1, extracting design variables affecting the separation trajectory of the aircraft and missile during separation; the design variables include first-type parameters, second-type parameters, and third-type parameters; the first-type parameters are parameters related to the initial state of the missile during separation, the second-type parameters are parameters related to the initial environment, and the third-type parameters are parameters related to missile body control; S2, based on the design variables in S1, selecting the parameters to be affected by the perturbation, and adding a small bias quantity with an artificially set distribution law to introduce the selected perturbation effect; the small bias quantity is a preset independent distribution function, or an actual value, or a distribution relationship measured by flight tests. The method for applying perturbation factors in missile-aircraft separation provided by this invention more closely reflects the real situation, providing designers with an effective method to examine the perturbation effects of parameters in the design.

The application discloses a fault probability monitoring and regulation system and method, relates to the technical field of fault monitoring, and collects to-be-determined related parameters in each period through a collection module, determines the optimal order of an autoregressive model of each related parameter by using the Akaike information criterion, and verifies the to-be-determined related parameters in each period, decides whether to perform secondary collection and replacement based on a verification result, generates a related parameter vector of each period, and starts a window translation along a period in a reverse order from the related parameter vector of each period through a monitoring module, calculates and arranges a comprehensive feature vector in each window, extracts and generates a core feature vector set through a time series self-encoder, trains a hidden Markov model by using a forward-backward algorithm, and obtains the fault probability of each period based on Bayesian inference, compares the fault probability with a probability threshold value to determine whether a fault exists, and executes regulation when the fault exists, so that high-precision fault monitoring and regulation based on collection self-checking, multi-dimensional data fusion and probability inference are realized.

The invention discloses a configuration updating method and a related device. The method comprises the following steps: acquiring monitoring data corresponding to a first model; whether configuration information of a first channel is updated or not is determined according to changes of the monitoring data, the first channel is used for transmitting all or part of first data, and the first data comprises at least one of the monitoring data, training data of a first model and reasoning data of the first model. By monitoring the model parameters, whether the change of the parameters corresponding to the model needs to cause updating of the channel configuration information for transmitting the model parameters or not is judged, so that the suitability of the configuration related parameters of the channel and the corresponding parameters of the model is ensured, and the effectiveness of model life cycle management is improved.

The invention discloses a federated learning model training method based on distributed random convex difference optimization, and belongs to the technical field of federated learning model training, and the method specifically comprises the following steps: S1, system initialization; s2, according to the federated machine learning task selected in the S1 and the machine learning model, constructing a federated learning model training-oriented distributed optimization model with a convex difference structure; s3, the central server designs a solution architecture composed of external iteration and internal iteration and initializes relevant parameters required by the iteration process; s4, the client side executes local model parameter updating in parallel; s5, the central server executes global model parameter updating; and S6, the central server executes algorithm termination condition judgment and outputs optimal global model parameters. According to the federal learning model training method based on distributed random convex difference optimization, convergence is faster and more stable, and better generalization and interpretability can be obtained under non-convex regularization.

The application belongs to the technical field of data cleaning, and discloses a sensor drift data correction method for a nonlinear system for fault diagnosis, which comprises the following steps: step one, system state space models under conditions that sensors exist and do not exist measurement drift are respectively established, and relevant parameter setting and initialization are performed; step two, based on the model one, the system equation is linearized through Taylor expansion, prior estimation of performance parameters is performed, and prior estimation of performance parameters at the current time and prior estimation error variance are obtained; step three, a judgment criterion of measurement drift is constructed based on the prior estimation of performance parameters, and whether the current sensor appears measurement drift is judged; step four, according to the judgment result of measurement drift in the last step, different strategies are adopted to perform posterior estimation of performance parameters; and step five, fault diagnosis of the system is performed based on the posterior estimation of performance parameters at the current time. The application solves the problem of sensor measurement drift estimation in a space limited scene.

The present invention relates to a method for quickly tracing the surge of a power plant induced draft fan, including: conducting a surge simulation test on a unit model with a divided surge source area to obtain n samples of multiple fault categories containing r relevant parameter data, adding fault labels to all samples to construct a surge fault library; selecting K key parameters beneficial to surge tracing from r relevant parameters, and dividing the samples into a training set and a validation set for each parameter combination; converting the time-domain data of the key parameters in each sample into a two-dimensional feature map of the spectrum; using the two-dimensional feature map as the input and the probability distribution of the fault categories as the output to conduct convolutional neural network training and validation; for the newly occurring surge of the unit, obtaining the two-dimensional feature map of the measured point data corresponding to the positions of K key parameters, inputting it into the network model to obtain the probability distribution of the fault categories, and the fault category with the highest probability is the surge tracing result. It can quickly lock the fault range and reduce the losses caused by surges, which is beneficial to the intelligent operation of the power plant.

The invention relates to the technical field of data processing, in particular to a loom performance optimization method and system based on data acquisition, and the method comprises the following steps: collecting and setting weft density, rotating speed, actual weft density and power consumption, integrating the weft density, the rotating speed, the actual weft density and the power consumption into a synchronous parameter set, calculating correlation strength among parameters, and screening high-correlation parameters to form a feature sequence; parameters are set as independent variables, results are set as dependent variables, a quadratic regression model is input to construct a prediction model, and an optimal parameter combination is obtained by combining process target iterative optimization. According to the method, a parameter set containing a set value and a result value is constructed by collecting multi-dimensional operation data of the loom, internal relation quantification among parameters is achieved, significant influence factors are extracted through correlation screening, output reliability is improved through a regression prediction mechanism, global optimization is achieved through performance evaluation and iterative optimization under target constraint, experience one-sidedness is reduced, and the method is suitable for large-scale popularization and application. Energy consumption and efficiency balance is achieved, the operation stability and the resource utilization rate are improved, and production continuity and low cost are guaranteed.

The present invention provides a data-driven control method and system for a ball-bar system, comprising: applying continuous excitation to the system multiple times, collecting corresponding response information, obtaining input and output data, and then obtaining an equivalent representation of the system state feedback; using an equivalent state space expression, based on Lyapunov stability theory, giving the design constraint conditions for the ball-bar system state feedback control to meet the asymptotic stability requirements; solving the design constraint conditions for the state feedback control system to meet the asymptotic stability requirements, and then solving the system's stable state feedback gain using the relevant parameters that meet the constraint conditions. The present invention solves the technical problems of the existing ball-bar system controller design technology, such as complex controller design, complex parameter adjustment, and poor control accuracy.

The present invention discloses a method and system for evaluating and optimizing a housing building structure, which relates to the technical field of architectural design. The method includes: establishing a finite element model of the housing building, obtaining relevant parameters, and setting them in the model to analyze the value range of the parameters. Based on the value range, candidate values are selected according to rules, a parameter grid is constructed, and it is divided into a preset number of sub-regions. Parameters are randomly selected from each sub-region as a reference combination, an optimization objective function is set to calculate the target value of the sub-region, and the target sub-region is determined. The grid division and reference parameter selection are repeated for the target sub-region, and the range is gradually narrowed until the parameter space is smaller than the preset threshold. Finally, using an optimization algorithm and the objective function, the optimal housing building parameter combination is determined within the final target sub-region. The present invention solves the problem in the prior art that it is impossible to balance efficiency and accuracy when optimizing the housing building structure.