A Class Of Intelligent Algorithms And The Integration Study With The Variational Optimization

Global optimization problem, especially intelligent optimization algorithms andvariational, is widely seen in the economic model, financial computing, network traffic,system control, biological engineering, environmental engineering, and so on. Thispaper studies particle swarm optimization and harmony search algorithm in theintelligent optimization algorithm, and the integration with variational optimizationproblem.First, this article proposed particle swarm optimization algorithm based on variablescale combining better local optimization ability of variable metric method with particleswarm algorithm, while it was applied to solve nonlinear equations. Then, the improvedalgorithm was applied to solve partial differential equation, improving the traditionalfinite difference method, and combining with the improved particle swarm algorithm tosolve partial differential equations and being validated by numerical examples. A largenumber of numerical experiments showed that the improved particle swarm algorithmnot only improved the convergence speed and accuracy, increased optimizationefficiency, but also demonstrated a strong applicability and robustness.Secondly, this article proposed the harmony search algorithm based on variablemetric that combining harmony search algorithm with the variable metric of the betterlocal search ability. Verified by numerical experiments the improved algorithmoptimization ability is strong, robustness is better, and the method is easy to implement.Once again, this article explored to transfer the differential equations into avariational problem, combinig the right margin method of solving differential equationswith the standard particle swarm algorithm, proposing a new optimization method. Atthe same time, it proposed to the new method of solve variational problems thatcombining particle swarm optimization with least-squares method. Being given toimprovement ideas, calculation process, through comparing simulation experiments, thetwo improved variational optimization algorithm have strong applicability.