Execution method and electronic device of meta-heuristic algorithm based on gpu parallel computing

Abstract

The present application relates to an execution method and electronic device of a meta-heuristic algorithm based on GPU parallel computing, and belongs to the technical field of optimization algorithms. The execution method of the present application includes allocating an independent graphics card memory space for each ant and decoy in the algorithm, and initializing parameters Pass it into the GPU; compare and evaluate the decoy position based on iterations in the GPU. When the number of iterations reaches the maximum number of iterations, copy the positional parameters of the decoy with the best position from the video memory of the graphics card to the memory, release the video memory space of the graphics card, and output the result ; wherein, in each iteration of the comparative evaluation of the bait position, including: parallel calculation of the objective function value of each ant and each bait about the position parameter, according to the comparison of the objective function value, update the position of the bait, and compare and determine the position The best bait; calculate the position of each ant parade after selecting the target bait in parallel, and update the position of the ant. The technical solution of the present application is beneficial to meet the real-time requirements in applications.

Description

technical field [0001] The present application belongs to the technical field of optimization algorithms, and in particular relates to an execution method and electronic device of a meta-heuristic algorithm based on GPU parallel computing. Background technique [0002] The general formulation of the optimization problem in engineering is to select a set of parameters (variables) to make the design index (objective) reach the optimal value under a series of related constraints (constraints). There are many optimization algorithms at present, such as gradient descent method, Newton method, conjugate gradient method, meta-heuristic optimization method, etc. [0003] The gradient descent method is simple to implement. When the objective function is a convex function, the solution of the gradient descent method is a global solution. But in general, the solution is not guaranteed to be the global optimal solution, and the gradient descent method is not the fastest. Newton's meth...

AI Technical Summary

Problems solved by technology

The disadvantage of the general meta-heuristic algorithm is that the calculation and solution speed is slow during the application process, and when facing a large-scale search problem, the ability to jump out of the local optimum is poor.

Images