274 results about "Surrogate model" patented technology

A method and system for performing multi-objective optimization of a multi-parameter design having several variables and performance metrics. The optimization objectives include the performance values of surrogate models of the performance metrics and the uncertainty in the surrogate models. The uncertainty is always maximized while the performance metrics can be maximized or minimized in accordance with the definitions of the respective performance metrics. Alternatively, one of the optimization objectives can be the value of a user-defined cost function of the multi-parameter design, the cost function depending from the performance metrics and / or the variables. In this case, the other objective is the uncertainty of the cost function, which is maximized. The multi-parameter designs include electrical circuit designs such as analog, mixed-signal, and custom digital circuits.

Methods for creating and using space-time surrogate models of subsurface regions, such as subsurface regions containing at least one hydrocarbon formation. The created surrogate models are explicit models that may be created from implicit models, such as computationally intensive full-physics models. The space-time surrogate models are parametric with respect to preselected variables, such as space, state, and / or design variables, while also indicating responsiveness of the preselected variables with respect to time. In some embodiments, the space-time surrogate model may be parametric with respect to preselected variables as well as to time. Methods for updating and evolving models of subsurface regions are also disclosed.

A method and system for personalized non-invasive assessment of renal artery stenosis for a patient is disclosed. Medical image data of a patient is received. Patient-specific renal arterial geometry of the patient is extracted from the medical image data. Features are extracted from the patient-specific renal arterial geometry of the patient. A hemodynamic index is computed for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model. The machine-learning based surrogate model is trained based on features extracted from synthetically generated renal arterial geometries.

A method of designing or selecting an implantable medical device comprises the steps of: i) obtaining a plurality of measured data points of a characteristic of an anatomical feature of an individual; ii) using said data points to construct a surrogate model of said characteristic, the surrogate model being constructed by interpolating or regressing measured data points of the characteristic, and using said surrogate model to obtain predicted values of said characteristic at a plurality of locations; iii) using said predicted values to determine or select at least one value of a design parameter of the implantable medical device. There is further disclosed a method of monitoring or diagnosing a disease or disorder.

The present disclosure describes the use of growth models and data driven models that are combined for quickly and efficiently modeling SAGD reservoir oil production. Growth function surrogate models are used for efficient and reliable reservoir modeling and production forecasting as opposed to CPU intensive simulations based on finite difference models. A data-driven technique can then compare the growth function surrogate model with real field data to find discrepancies and inconsistencies between the two, allowing for an updates and improvements of the growth function model.

The invention provides an optimization design method based on a self-adaptive radial basis function surrogate model for an aircraft. The optimization design method comprises the following steps of: first, sampling an experimental design sample in a design space by adopting a latin square experimental design method and acquiring an aircraft high-precision analytical model response value corresponding to the experimental design sample; constructing an approximate aircraft high-precision analytic model of the radial basis function surrogate model; acquiring the global optimal solution of a current radial basis function surrogate model by utilizing a genetic algorithm; constructing an aircraft optimization design major sampling space according to current optimization flow information, increasing a few experimental design samples, and updating the radial basis function surrogate model; and acquiring the global optimal solution of the updated radial basis function surrogate model by utilizing the genetic algorithm again, judging whether an optimization flow is converged or not, stopping optimization if the optimization flow is converged, and reconstructing the aircraft optimization design major sampling space until the optimization is converged if the optimization flow is not converged. By using the optimization design method provided by the invention, the optimization efficiency is improved, and the optimization design cost of the aircraft is saved.

The invention relates to a POD (proper orthogonal decomposition) and surrogate model based order reduction method for hypersonic aerodynamic thermal models and belongs to the technical field of aerospace. A hypersonic aircraft aerodynamic thermal environment is predicted and obtained by the aid of a POD and surrogate model method, nonlinear characteristics of real gas effect, turbulence viscosity and the like of high-precision numerical calculation are reserved, the hypersonic aircraft aerodynamic thermal environment can be provided for design of the hypersonic aerocraft by the aid of model order reduction method which is high in precision and efficiency, related thermal boundary conditions are provided for aerodynamic thermal elastic design, the thermal environment is provided for thermal protection design of the hypersonic aerocraft, design efficiency can be greatly improved, design circle is shortened, and design cost is reduced.

A method of optimising a sequential combinatorial process comprising an interchangeable sequence of events comprises using a master model to model a selection of the possible sequences, and using information derived from the master model in a surrogate model that approximates the master model with a much shorter computation time. The surrogate model calculates all the possible sequences using an algorithm to select from the information calculated by the master model that which most closely matches the events of a present sequence, following a prioritised system so that the best match is used wherever possible. All results from the surrogate model are compared so that the modelled sequence that gives the result closest to a desired optimum result for the process can be identified, and potentially applied to the process. Accuracy can be enhanced y running the optimum sequence through the master model as a check, and further by adding the optimum sequence to the information used by the surrogate model in future calculations. Any sequential combinatorial process, in which the quality of the end result of the process depends on the order in which events in the process are performed, can be optimised in this way, including manufacturing process such as machining, cutting, shaping, forming and / or heat treating a workpiece, flow of material through a factory or oil or gas through a pipeline network, chemical and material science mixing processes, computational biology modelling, and fleet management.

Systems and methods which implement surrogate (e.g., approximation) models to systematically reduce the parameter space in an optimization problem are shown. In certain embodiments, rigorous (e.g., higher fidelity) models are implemented with respect to the reduced parameter space provided by use of surrogate models to efficiently and more rapidly arrive at an optimized solution. Accordingly, certain embodiments build surrogate models of an actual simulation, and systematically reduce the number of design parameters used in the actual simulation to solve optimization problems using the actual simulation. A multi-stage method that facilitates optimization of decisions related to development planning and reservoir management may be provided. Iterative processing may be implemented with respect to a multi-stage optimization method. There may be uncertainty in various parameters, such as in reservoir parameters, which is taken into account according to certain embodiments.

A method and system for estimating tissue parameters of a computational model of organ function and their uncertainty due to model assumptions, data noise and optimization limitations is disclosed. As applied to a cardiac use-case, a patient-specific anatomical heart model is generated from medical image data of a patient. A patient-specific computational heart model is generated based on the patient-specific anatomical heart model. Patient-specific parameters and corresponding uncertainty values are estimated for at least a subset of parameters of the patient-specific computational heart model. A surrogate model is estimated for a forward model of cardiac function, and the surrogate model is applied within Bayesian inference to estimate the posterior probability density function of the parameter space of the forward model. Cardiac function for the patient is simulated using the patient-specific computational heart model. The estimated parameters, their uncertainty, and the computed cardiac function are displayed to the user.

The invention relates to an optimum design method for an aircraft system based on a sequence radial basis function surrogate model and belongs to the technical field of optimum design of the aircraft system. According to the optimum design method, the thought of searching step range of trust domain solving high-dimensional optimization problem is applied to a supervisory sequence surrogate model, the sample space is gradually updated and the approximate precision of the surrogate model is promoted, so that the optimizing strategy convergence is guided to the globally optimal solution and the optimum design method provided by the invention has better global optimization capacity and optimization efficiency. The method is strong in universality. According to the method, the program development is conveniently realized, the optimum design means for the aircraft system is improved, the design cost is lowered and the requirement for multidisciplinary optimum design of the present aircraft system is met.

The invention discloses a test run method for an accelerated life test of a gas turbine engine. The test run method comprises the steps that an accelerated factor analysis model of the gas turbine engine under simulated running conditions is established; according to the accelerated factor analysis model, a support vector machine method is used for establishing a surrogate model of three-dimensional finite element stress analysis; the surrogate model is used for obtaining dangerous point stress of the accelerated factor analysis model, and then life indexes of the accelerated factor analysis model are obtained; a Monte Carlo method is used for conducting stochastic simulation on random variables of the life indexes; an optimal design mathematical model of an accelerated life test scheme of stress under the simulated running conditions of the gas turbine engine is established and analyzed, and the test scheme is optimized according to the optimal design mathematical model; a mixing optimization method is adopted to optimize the mathematical model, and the optimized accelerated life test scheme is obtained according to the optimized mathematical model. By means of the test run method, the test period is shortened, and life test expenditure is reduced.

A system includes a schedule generator having one or more processors configured to obtain resource extraction parameters for extracting a resource from a reservoir. The resource extraction parameters include well creation parameters associated with drilling wellbores, well stimulation parameters associated with introducing fracturing fluid into the wellbores, and production parameters associated with extracting the resource through the wellbores. The schedule generator selects initial trial schedules having different values of the resource extraction parameters and receives initial resource output data generated by execution of the initial trial schedules with a designated reservoir model. The schedule generator generates a surrogate model based on the initial resource output data and the initial trial schedules and uses the surrogate model to perform iterations of selecting modified trial schedules until a predetermined condition is satisfied.

Embodiments provide methods and systems for constructing surrogate models for use in interactive experiences. One such embodiment begins by defining a model that includes a parametric state vector and a design variable vector and represents a real world system. Next, a first and second experiment are performed to determine a response over time of the parametric state vector and to produce a dataset of the parametric state vector and the design variable vector as a function of time. The dataset is then modified with one or more derivatives of the parametric state vector and a set of surrogate differential equations is constructed that approximates a higher derivative of the parametric state vector relative to that in the dataset and the set of surrogate differential equations is stored as a surrogate model. The surrogate model is in turn provided from memory in a manner accelerating simulated behavior in response to user-interaction with the model.

The invention provides a helicopter moving part service life management method based on digital twinning. The helicopter moving part service life management method comprises the steps of load data collection, structural stress analysis, crack parametric modeling, fracture mechanics simulation, agent model construction of fracture mechanics parameter prediction, probabilistic prediction of crack propagation, structural risk evaluation and dynamic adjustment of a maintenance inspection plan. The invention relates to an electronic device and a storage medium, which are used for executing a helicopter moving part life management method based on digital twinning. According to the invention, multiple scales of the structure are considered and multiple models are fused; and establishing a digitaltwinning body of the typical moving part structure of the helicopter, performing high-performance real-time simulation according to the flight parameters of the helicopter and the load and damage data of the key part of the structure, dynamically updating the damage state of the typical moving part of the helicopter, and adjusting the residual life and the maintenance and inspection plan of the structure.

The invention relates to the technical field of stability checking of main strength-bearing thin-walled members of aerospace and architectural structures, and discloses a method for determining a reduction factor of a bearing capacity of an axial load cylindrical shell structure, which is different from an experiment experience-based conventional defect sensitivity evaluating method with NASASP-8007 as a representative. Pit defect is introduced in a mode of applying a radial concentrated force (disturbance load). The method comprises the steps of: firstly, carrying out numerical analysis on an influence law of pit defect amplitude of a single point to the axial load bearing capacity of the axial load cylindrical shell shaft, and determining a reasonable load amplitude range; secondly, carrying out defect sensitivity analysis on pit defects of multiple points; thirdly, carrying out experiment design sampling by using load amplitude values and loading position distribution as design variables; and finally, based on optimizing technologies such as an enumeration method, a genetic algorithm and a surrogate model, searching the most disadvantageous disturbance loads limiting the defect amplitude in multiple points, and determining the reduction factor of the bearing capacity of the axial load cylindrical shell structure, so as to establish the more actual and reliable method for evaluating the defect sensitivity and the bearing performance of the axial load cylindrical shell structure, wherein the method has great physical significance.

The invention discloses a real-time yield prediction method for a hydrocracking device. Field real-time data are processed with a data reconciliation technology, and hydrocracking reaction kinetics parameters are corrected in real time in combination with an improved differential evolution algorithm, so that a mechanism model can accurately describe the actual running condition of the device. On the basis of the corrected model, effects caused by key operation / process conditions such as the raw material density, the sulfur content, the nitrogen content, the reaction temperature, the pressure, the hydrogen-to-oil volume ratio and the like on hydrocracked products are analyzed. Piecewise linearization is performed according to the effect trend, a linear equation is solved, corresponding Delta-Base yield data are acquired, the operation condition is associated with the Delta-Base data with a neutral network modeling technology, a yield surrogate model is established, the yield data calculation speed is increased, real-time prediction of the yield of products of the hydrocracking device is realized, and theoretical support is provided for establishing an accurate plan optimization PIMS (process industry modeling system) model.

The invention relates to the technical field of bearing designing, in particular to a multi-objective optimization design method of a spiral oil wedge bearing. According to the method, an oil film property calculation method based on solving a Navier-Stokes function is used for calculating the oil film properties of the spiral oil wedge bearing, a minitype multi-objective genetic algorithm is used for conducting multi-objective optimization design of a bearing structure, and a radial basis function surrogate model is established in the optimization process for calculating. The method comprises the following specific steps of establishing a mathematical model of a spiral oil wedge bearing multi-objective optimization design problem, sampling experimental design methods, establishing the radial basis function surrogate model, evaluating the precision of the surrogate model, using the minitype multi-objective genetic algorithm for solving the multi-objective optimization design problem and acquiring an optimal compromise solution. The multi-objective optimization design method is suitable for the multi-objective project optimization problem of the spiral oil wedge bearing, high global optimization capacity is achieved, the optimization efficiency can be effectively improved, the solving efficiency and the calculation precision of the algorithm are improved, and the time for calculation is little.

A method and system for determining hemodynamic indices, such as fractional flow reserve (FFR), for a location of interest in a coronary artery of a patient are disclosed. Medical image data of a patient is received. Patient-specific coronary arterial tree geometry of the patient is extracted from the medical image data. Geometric features are extracted from the patient-specific coronary arterial tree geometry of the patient. A hemodynamic index, such as FFR, is computed for a location of interest in the patient-specific coronary arterial tree based on the extracted geometric features using a trained machine-learning based surrogate model. The machine-learning based surrogate model is trained based on geometric features extracted from synthetically generated coronary arterial tree geometries.

The invention discloses an optimizing method for a multi-flexible dynamic structure of a bridge crane. A multi-flexible dynamic technology, an optimal Latin hypercube algorithm, a neural network model optimized through particle swarm optimization, a NSGA-II algorithm and Maxi-min criterion are adopted for solving the problem of difficulty in optimizing the flexible part in the previous multi-flexible dynamic optimization process. According to the method, design variable values in a dynamic model are changed by altering a finite element parameter for changing modal neutral file information; a BP neural network optimized by the particle swarm optimization is introduced for establishing a surrogate model; the non-linear relationship between the design variable values of the flexible body and the optimized target value in the multi-flexible dynamic model is fit; the NSGA-II genetic algorithm is adopted for performing multi-target optimization on the surrogate model, thereby acquiring a Pareto solution set; the Maxi-min criterion is adopted for finding a feasible solution considering all the optimized targets.

The invention relates to a complete period dynamic optimization method for an industrial ethylene cracking furnace and based on a surrogate model. By simulating the mechanism of the industrial ethylene cracking furnace and utilizing an experimental design principle, a certain amount of simulation data is generated, then a neural network surrogate model is utilized to carry out modeling. By employing the neural network surrogate model and combining iterative computation update of the thickness of a coking layer in a furnace tube, a complete period dynamic model of the cracking furnace is obtained. Based on the complete period dynamic model of the cracking furnace, the complete period dynamic optimization method for the cracking furnace is provided, then a constructed continuous dynamic optimization model is approximately solved by means of a piecewise linear representation method. Compared with experiential operation, dynamic optimization is substantially greater in economic benefit. The complete period dynamic optimization method is simple in theory, reasonable in derivation, high in engineering usability, easy to transplant and wide in adaptability, and is simple and practicable.

The invention discloses a parameter optimization method of Modelica model based on surrogate model, which comprises the following steps: 1, compiling Modelica model and obtaining model parameter and variable information; 2, optimizing modeling; 3, generating sampling point; 4, carry out simulation calculation on that parameter combination; 5, analyzing that simulation calculation result; 6, constructing a proxy model; 7, using that surrogate model to replace the Modelica model to carry on the optimization iteration and find the optimal parameter; 8, carrying out simulation calculation on thatoptimal parameters, and if the error between the simulation calculation result and the output result of the proxy model is small than a set value, executing the step 10, otherwise executing the step 9; 9, dynamically updating the agent model according to the simulation calculation result of the step 8, and then executing the step 7; 10, the optimal parameter calculated in the step 7 is the final optimization result, and the parameter optimization is finished; Through the above steps, the invention achieves the purpose of improving the parameter optimization efficiency of the Modelica model, and solves the practical problem that the calculation amount in the parameter optimization process of the Modelica model is huge and it is difficult to optimize the parameters of the large-scale model.

The invention discloses a design method of a wing type of a rotor wing of a micro-miniature type rotor wing UAV and a product, and the design method combines an efficient optimization algorithm basedon the agent model, a shape parameterization method based on the free deformation technique, a grid deformation method based on the radial basis function and an aerodynamic performance computing method based on the average Reynolds equation. Firstly, the design condition is determined, the incoming flow states, namely, the incoming flow Mach number, the Reynolds number and the angle of incidence,of the wing type are determined according to the main operation condition of a propeller and the radial position of the wing type on the blades; then the design index is determined; a lift-drag ratioas large as possible is obtained. According to the method, an Eppler 387 is taken as the datum wing type, the wing type whose whole curvature of the upper surface protrudes upward, front section of the curvature of the lower surface is concaved downward and rear section protrudes upward is designed, the relative thickness is small, the upper surface of the rear section is relatively flat, the whole curvature is large, the front section of the lower surface is concaved remarkably, the wing type has low total resistance coefficient and high lift coefficient, and accordingly the wing type has high lift-drag ratio and the aerodynamic efficiency is improved.

The invention discloses a full-process modeling method for an oil refining process. Based on the mechanisms and running characteristics of all production devices in the oil refining process and a corrected model, the influence of key operation / process conditions of all the devices on the product yield is analyzed. Piecewise linearization is carried out according to the influence trend, a linear equation is solved, corresponding Delta-Base yield data are obtained, a neural network modeling technology is combined, operation conditions and the Delta-Base data are related, a yield surrogate model is built, the yield data calculating speed is improved, real-time prediction on the product yield in the oil refining process is achieved, and theoretical supports are provided for building a precise plan optimized PIMS model.

A method and system for performing multi-objective optimization of a multi-parameter design having several variables and performance metrics. The optimization objectives include the performance values of surrogate models of the performance metrics and the uncertainty in the surrogate models. The uncertainty is always maximized while the performance metrics can be maximized or minimized in accordance with the definitions of the respective performance metrics. Alternatively, one of the optimization objectives can be the value of a user-defined cost function of the multi-parameter design, the cost function depending from the performance metrics and / or the variables. In this case, the other objective is the uncertainty of the cost function, which is maximized. The multi-parameter designs include electrical circuit designs such as analog, mixed-signal, and custom digital circuits.

A method for determining a reduction factor of a bearing capacity of an axial load cylindrical shell structure relates to stability checking of main bearing strength thin-walled members of aerospace and architectural structures. Different from experiment experience-based conventional defect sensitivity evaluating method represented by NASA SP-8007, a depression defect is introduced in a manner of applying a radial disturbance load. First, an influence rule of a depression defect amplitude of a single point to an axial load bearing capacity is analyzed by using numerical values, so as to determine a load amplitude range; then, defect sensitivity analysis is performed on depression defects of multiple points; then, experiment design sampling is performed by using load amplitude values and load position distribution as design variables; and finally, based on optimizing technologies such as an enumeration method, a genetic algorithm and a surrogate model, the most disadvantageous disturbance load of the multiple points that limits the defect amplitude is searched for, and a reduction factor of the bearing capacity of the axial load cylindrical shell structure is determined, so as to establish a more physical method for evaluating the defect sensitivity and the bearing performance of the axial load cylindrical shell structure.

The invention discloses an iterative design method for a hypersonic-velocity aircraft model. The method comprises the steps that a parameterized model of a hypersonic-velocity aircraft is built through a geometric design method; a key model parameter of the hypersonic-velocity aircraft is extracted through a sensitivity approach, the parameterized model is simplified, and a surrogate model of the hypersonic-velocity aircraft for the iterative design is obtained; the model parameter of the hypersonic-velocity aircraft is optimized by using a pigeon-inspired optimization algorithm, and the balance states of the model in different flight conditions are obtained; a desired performance index of the hypersonic-velocity aircraft is determined, the balance states of the model are subjected to iteration, and the optimal design model of the aircraft is obtained. In complex flight conditions, the optimal design model is obtained by using the pigeon-inspired optimization algorithm according to the desired performance index, the optimal model parameter meeting the desired performance index is quickly obtained, all flight envelopes can be covered, and a good model optimizing tool is provided for the design of the hypersonic-velocity aircraft.

A computer-implemented method for personalized assessment of a subject's bone health includes extracting a plurality of features of interest from non-invasive subject data, medical images of the subject, and subject-specific bone turnover marker values. A surrogate model and the plurality of features of interest are used to predict one or more subject-specific measures of interest related to bonehealth. Then, a visualization of the one or more subject-specific measures of interest related to bone health is generated.

The invention provides an electromechanical cooperative design method of a skin antenna based on parallel Bayesian optimization in order to reduce calculation amount, improve design quality and optimize efficiency. Determine the design variables and initial design space, carry out initial sampling, extract all sample points in the sample point database and their corresponding response values, select Bayesian algorithm to construct the surrogate model of the objective function, use the pseudo-expectation enhancement criterion as the collection function to select multiple candidate solutions, and judge the degree of satisfaction of the constraint conditions. If the number of calculation times of the electromechanical coupling model reaches the set number or satisfies the convergence criterion, the iteration is stopped and the optimal solution is outputted. A large amount of optimization time is saved for complex problems, and only one sample point is randomly selected for initial sampling, so the optimization process is repeatable.