US 2026/0228300 A1 · Method for Solving System of Nonlinear Equations and Related Device

Abstract

Embodiments of this application disclose a method for solving a system of nonlinear equations. In the method, during an ith iteration of a first system of nonlinear equations, a neural network model can identify, based on input description information of the ith iteration, system of equations feature information of the first system of nonlinear equations during the ith iteration, to efficiently determine a first convergence policy that is appropriate for the first system of nonlinear equations in the current iteration process. In this way, a target convergence policy is determined according to the first convergence policy, and the first system of nonlinear equations is iteratively solved efficiently and accurately according to the target convergence policy, so that performance of iterative solving on a system of nonlinear equations is improved (for example, a solving speed is improved and solving time consumption is reduced).

Claims20

Dependent claims

A method for solving a system of nonlinear equations, applied to a compute device cluster, wherein the method comprises:

obtaining a first system of nonlinear equations;
obtaining, based on the first system of nonlinear equations, description information of an ith iteration of the first system of nonlinear equations, wherein the description information of the ith iteration is used to describe system of equations feature information of the first system of nonlinear equations during the ith iteration, and i is a positive integer;
performing inference on the description information of the ith iteration of the first system of nonlinear equations through a neural network model, to obtain at least one first convergence policy of the ith iteration;
and
determining a target convergence policy according to the at least one first convergence policy, and solving the first system of nonlinear equations through the ith iteration according to the target convergence policy, to obtain a target solution of the first system of nonlinear equations.

2 more elements

An apparatus for solving a system of nonlinear equations, wherein the apparatus comprising a processor, a memory, wherein the memory is configured to store an instruction, and the processor is configured to invoke the instruction in the memory to:

obtain a first system of nonlinear equations;
and
obtain, based on the first system of nonlinear equations, description information of an ith iteration of the first system of nonlinear equations, wherein the description information of the ith iteration is used to describe system of equations feature information of the first system of nonlinear equations during the ith iteration, and i is a positive integer;
perform inference on the description information of the ith iteration of the first system of nonlinear equations through a neural network model, to obtain at least one first convergence policy of the ith iteration;
and
determine a target convergence policy according to the at least one first convergence policy, and solve the first system of nonlinear equations through the ith iteration according to the target convergence policy, to obtain a target solution of the first system of nonlinear equations.

3 more elements

Dependencies are inferred from the claim text.

Method for Solving System of Nonlinear Equations and Related Device

Life of this patent

Method for Solving System of Nonlinear Equations and Related Device

Life of this patent

Abstract

Claims20