14 results about "Functional relation" patented technology

Methods, systems, and computer-readable media are used in graph-based machine learning of analog integrated circuits (ICs). In a first aspect, functional device pairs in transistor-level circuits are detected and classified using a hybrid approach combining a subgraph isomorphism algorithm, such as VF2, with a trained relational-graph convolutional network (RGCN). The VF2 algorithm identifies candidate pairs and initial categories, while the RGCN filters false positives using link prediction scores, preserving category labels for validated pairs. In a second aspect, a relational GraphSAGE model performs multi-class link prediction on a heterogeneous graph with netlist and functional relation edge types, labeling device pairs into analog primitive categories without technology-dependent features and merging overlapping pairs into larger functional groups. In a third aspect, circuit performance is predicted using a hierarchy-aware graph neural network comprising an edge-conditioned convolution (ECC) layer and multiple Circuit graph isomorphism network layers corresponding to hierarchy levels of device groupings.

The invention relates to a computer-implemented method (100) for solving at least one differential equation of at least one physical model, comprising at least the following steps: training (102) at least one neural operator (10) using input training data and output training data associated with the at least one differential equation, wherein the at least one neural operator (10) performs at least one Fourier transform using at least one learning quantum Fourier transform quantum circuit (20), wherein the neural operator (10) and the learning quantum Fourier transform quantum circuit (20) are trained such that the at least one neural operator (10) outputs at least one functional relationship between the input training data and the output training data;and determining (104) at least one solution (28) of the at least one differential equation using the at least one trained neural operator (10). The method (100) exhibits increased time efficiency.;

This invention discloses a method, system, and equipment for disaster relief using swarm robots. The method includes: simulating a disaster relief scenario; determining the hard rules that each robot in the swarm must follow based on its actions and states in the simulated scenario; constructing a functional relationship between the adoption rate of social rules and the performance indicators of the swarm robots; constructing a compromise decision model based on the performance indicators, the social rules, and the functional relationship; solving the compromise decision model using an adaptive linear programming algorithm and performing multi-objective visual trade-off analysis to obtain the range of social rule adoption rates that satisfies all the performance indicators; and directing the swarm robots to perform disaster relief based on the range of social rule adoption rates. This invention can minimize casualties and property damage in disaster relief.

This invention provides a method and system for optimizing feed rate in high-efficiency constant-force cutting with complex trajectories. The method includes: establishing a functional relationship model between spindle torque and feed rate; constructing a multi-objective optimization function based on the functional relationship model, with the objectives of minimizing machining time and minimizing spindle torque fluctuation; solving the multi-objective optimization function under preset constraints to obtain an optimal solution set; selecting the optimal solution from the optimal solution set to generate an optimized feed rate sequence, and updating the CNC machining code accordingly. This invention establishes a direct functional model between feed rate and spindle torque, constructs a multi-objective optimization problem with the core objectives of minimizing machining time and minimizing torque fluctuation, and introduces constant torque constraints, feed rate limits, and smoothness constraints. It employs a hybrid intelligent algorithm combining NSGA-II and Particle Swarm Optimization (PSO) for efficient solution, thereby significantly improving machining efficiency while ensuring stable machining load.

The invention discloses an oversized transmission drum body structure optimization design method and system, and the method comprises the steps: firstly building a drum body parameterized model in a finite element environment, building design variables of all structural parameters of a drum body, carrying out the grid division, adding constraints, and carrying out the static simulation, thereby obtaining the deformation of the drum body and all stress values of a dangerous section; creating and generating output variables according to the parameters; secondly, obtaining corresponding experimental data matrix data between multiple groups of structural design variables and output variables by utilizing an experimental design module in ANSYS; thirdly, importing experimental matrix data into Matlab, fitting to obtain a function relational expression between a structural design variable and an output variable, then creating a variable, a constraint function and an objective function of an optimization design model, creating an optimization program based on a heritage algorithm, and realizing the volume optimization of the cylinder on the premise of meeting various constraint conditions; therefore, the optimization design of the cylinder structure is completed; according to the scheme, by means of the method, the precision of optimization design of the oversized transmission drum body structure is further improved.

The invention provides an optimization method and system of a hydraulic servo valve and a storage medium. The optimization method of the hydraulic servo valve specifically comprises the following steps that boundary conditions and constraint conditions of the hydraulic servo valve are obtained, and a mathematical model is established; an initial three-dimensional model is established, the working state of the initial three-dimensional model is simulated, and a simulation result is obtained; determining an optimization variable, obtaining a preset optimization target, and establishing a function relation between the optimization variable and the preset optimization target; performing iterative mating on the optimization variable by using a genetic algorithm, continuously obtaining a mating result, and stopping iteration if a preset condition is met; and screening an iteration result, outputting an optimal value of the optimization variable, and constructing an optimal three-dimensional model. And establishing a three-dimensional model and performing simulation to obtain a function relationship, finding an optimal value of an optimization variable through a genetic algorithm, obtaining an optimal structure of the servo valve, and efficiently obtaining the optimal structure of the hydraulic servo valve.

The invention provides a passenger vehicle pedal parameter optimization design method and system based on AR virtual verification, and relates to the technical field of whole vehicle optimization layout, and the method comprises the steps: setting a standard coordinate system of a driver seat; setting a distance arrangement range between the Y direction and the center line of the driver in the standard coordinate system; according to a set distance arrangement range, fitting a function relationship between the center point of the initial position of the Z-direction pedal of the coordinate system and the height and the seat height of the AHP point; based on the function relation between the center point of the initial position of the Z-direction pedal and the height and the seat height of the AHP point, the function relation between the pedal surface angle and the seat height is further fitted, and arrangement design of the pedal is completed; and carrying out AR virtual verification on the designed pedal arrangement, advancing an evaluated time node from an engineering sample vehicle stage to a design stage, and identifying and determining an optimal pedal arrangement scheme. The problem of relevance between the pedal and the human body is solved, and the time cost of arrangement design is reduced.

The present invention relates to the technical field of bridge engineering, and particularly relates to a parametric design method for a multi-pylon extradosed bridge for a high-speed railway. On the basis of a provided main span length sequence, a functional relational expression of related design parameters is established to obtain rational schemes for a side-to-main span ratio, a height-to-span ratio and a pylon-to-span ratio; and comprehensive comparison and selection are performed on the basis of a main bridge cost calculation equation and the technical difficulty of engineering, and the methods of normalization and weighted averaging are used to perform data processing, so as to determine optimal schemes. The advantages of the present invention lie in: a theoretical basis is provided during an overall scheme design, and subjective human decision-making is reduced, thereby improving the refinement degree and rationality of the design.

This invention provides a method and apparatus for measuring magnetic flux density using the resistivity of molten metal. The method includes: establishing a functional relationship model between the resistivity and magnetic flux density of a target molten metal at different temperatures; measuring the resistivity of the target molten metal in real time after heating it to a molten state in a high-temperature magnetic field environment; acquiring the measurement temperature of the target molten metal in real time using a temperature sensor in the same high-temperature magnetic field environment; the high-temperature magnetic field environment being a magnetic flux density range of 2~6T and an ambient temperature of 250~1500℃; and inputting the measured resistivity and measurement temperature into the functional relationship model to obtain the magnetic flux density corresponding to the high-temperature magnetic field. This invention enables direct measurement of magnetic flux density under extreme temperature and magnetic field conditions.

A kind of modular operating arm collaborative driving control method for multiple linear drives belongs to mechanism system driving control field, it is characterized in that: analysis modular operating arm driving system, obtains the number of linear driver and actuation parameter;Select linear driver driving control parameter;Drive single linear driver to judge whether controllable actuation can be realized;If so, selected linear driver driving control parameter is used as the driving control parameter of all linear drivers;Set lower computer control logic;Planning trajectory;Establish operating arm kinematics model, obtain the functional relationship of driving control parameter and time;According to the functional relationship, the trajectory is converted into upper computer driving control program instruction to drive multiple linear drivers, complete preset trajectory task.This method can collaboratively drive and control multiple linear drivers of modular operating arm, greatly simplify the control method and process, and can be applied to the driving control of modular operating arm with multiple linear drives, and has strong versatility.

The invention relates to the technical field of metal material rolling, and provides a plate blank size optimization method and system for a wide and thick plate hot rolling process. According to the method, the shape change rule of the head, the tail and the edges after rolling is quantitatively described by constructing a yield prediction model of a rolled piece based on a plane shape model; establishing a function relationship by taking a broadening ratio and an elongation ratio as optimization variables and minimizing the loss amount of trimming and head and tail cutting as a target; and under the constraint of rolling mill equipment capacity, global optimization is carried out on the process parameter combination by adopting an inversion algorithm, and finally, the optimal slab size is derived through inversion. According to the method, conversion from experience judgment to scientific calculation is achieved, the yield potential is explored by precisely controlling the cutting loss source, the precision and efficiency of blank design are remarkably improved, and efficient and reliable decision support is provided for hot rolling production of wide and thick plates.

A device and method for analyzing a technical system or process that includes a first function and a second function. A first model models a functional relationship of input and output variables of the first function and an influence of a first disturbance variable. The second model models a functional relationship of input and output variables of the second function and an influence of a second disturbance variable. An observation is provided that includes input variables and output variables of the first function and of the second function. A prediction regarding an effect of changing the first function under the influence of the first disturbance variable and / or changing the second function under the influence of the second disturbance variable is determined, and of these functions, that function whose change has a higher likelihood of achieving a desired effect is selected.

The application discloses a robot polishing and grinding force hybrid control method and system for a large-curvature complex structure. The method comprises the following steps: establishing a functional relationship among the robot end contact force, the soft device output force, the friction force, and the end load by using Newton-Euler equation, and establishing a relationship model related to the material removal rate and a relationship model between the grinding curvature and the ideal polishing force by using Hertz contact theory; establishing a geometric relationship model among the robot end posture, the soft device inclination, and the workpiece surface curvature by using Euler angle method; introducing a PID algorithm to adjust the pose of the robot end adhesion state; adopting different force control modes for different stages of robot operation, real-time controlling the polishing process, and optimizing the robot operation performance. The application can comprehensively optimize the robot feed speed, the process parameters, and the end soft device hybrid control in the robot polishing and grinding process, so as to realize small-amount precise removal of the large-curvature complex structure.

The invention relates to a test method for obtaining axial deformation of a gas turbine casing based on local strain of a front support, and aims to solve the existing problems of high test cost and large calculation error of an axial deformation experiment of the gas turbine casing. According to the scheme, the maximum main strain of the central areas of the air inlet side bottom and the exhaust side top of the front support under different axial displacements of the gas turbine casing is obtained through finite element calculation, a function relation expression of the maximum main strain and the maximum main strain is obtained through fitting, and strain gauges are arranged in the central areas of the air inlet side bottom and the exhaust side top of the front support of the gas turbine respectively. Recording the maximum main strain of the central areas of the bottom of the gas inlet side and the top of the gas exhaust side of the front support of the gas turbine in real time under a slow working condition, a rated working condition or a maximum power working condition by adopting a strain tester, and substituting strain data measured by the strain tester in real time into the function relational expression; and finally, the axial deformation of the casing of the gas turbine under the slow working condition, the rated working condition or the maximum power working condition is obtained in real time, and the axial deformation of the casing is simply, quickly and accurately obtained in real time.