Homotopic neural optimizationtheory and its algorithm described in this paper can make nonlinear multiextremalobject functions converge fast to global extremum, so it is an effective inversionmethod.

The low discrepancy of quasi_random sequence ensures that the function field be searched evenly and various local extrema including global extremum be found.

The other is that it enlarges or shrinks fitness value properly,which,combined with random rule,determines a certain particle as global extreme value in speed update formula.

The characteristics of the optimization method based on the approximation technology, the traditional gradient-based search algorithm and the standard genetic algorithm are studied, through calculating global extreme of two test functions and re-constructing the geometry of airfoil.

The other is that it enlarges or shrinks fitness value properly,which,combined with random rule,determines a certain particle as global extreme value in speed update formula.

In this paper,a new theorem of global maximum and global minimum on finite dimensional Hilbert space was proposed and some applications were presented.

The reference frames to be estimated are analyzed first, the vectors accumulation algorithm is proposed based on the hypothesis that the error plane is monotone around the global match point in a small scale.

And the global best and personal best of particle swarm algorithm are selected by the grey relevancy theory. Further more,the grey particle swarm algorithm is presented for multi-objective models solution in the reliability robust optimization design.

The mutation operator was introduced by control model of discrete traffic signal to update the best point of the individual and the best point of the global in adaptive-Mutation particle swarm optimization algorithm (AMPSOA).

The global extremum (minimum) of this dependence, which was accepted as the goal function, was used as a criterion for designing the algorithm optimizing the gripper position on the load axis upon stabilizing the angular motion of the module.

Computer experiment demonstrated, however, that the iterative procedure stabilized after several steps, although the global extremum was not necessarily attained.

An effective method for finding the global extremum in identifying a logical-probabilistic risk model within reasonable computation time is developed.

Determination of the Nash-equilibria in the bimatrix game was considered by reducing it to its equivalent nonconvex problem of optimization solved by the algorithm of global search based on the theory of global extremum for this problem.

Parallelization of the global extremum searching process

It has not been adopted widely, partly because it requires global extrema, and not local and this has been regarded as a problem with no solution.

Conditions for coincidence of local and global extrema in optimization problems

Conditions for the coincidence of local and global extrema in problems of discrete optimization

This article presents a new algorithm, called the''Hyperbell Algorithm'', that searches for the global extrema ofnumerical functions of numerical variables.

Results show that if the global extrema of the nonconvex quadratic function are located on the boundary of the primal feasible space, the dual solutions should be interior points of the dual feasible set, which can be solved by deterministic methods.

The genetic algorithm is then employed to find out the global extreme solution of the cost function.

The genetic algorithm is then employed to find out the global extreme solution for the cost function.

Besides, the genetic algorithm will converge to global extreme instead of local extreme and achieves a good antenna pattern.

Finding coverage with minimum cost is an important step in logic minimization. Since it is a Np-complete problem, it is not practical for large input data. In this article, we use a stochastic neural network to find the optimal coverage through formalizing the coverage problem; choosing appropriate annealing schedule and parameters. The algorithm we use has low time and space complexity, and has high parallelism.

This correspondence states the criterion of cluster validity problem which injects validity measurement functionals into the final judgement i such as fuzzy partition coefficient and entropy.It involves the optimizing of validity functionals through the implementation of FCM algorithm to evaluate optimal clustering.In this way,a unique global minimum can be guaranteed for validity functional over fuzzy c-partition space.A clustering experiment for recognition of an ambiguous photograph of MIG- 29 fighter taken...

This correspondence states the criterion of cluster validity problem which injects validity measurement functionals into the final judgement i such as fuzzy partition coefficient and entropy.It involves the optimizing of validity functionals through the implementation of FCM algorithm to evaluate optimal clustering.In this way,a unique global minimum can be guaranteed for validity functional over fuzzy c-partition space.A clustering experiment for recognition of an ambiguous photograph of MIG- 29 fighter taken by a spy satellite is successfully developed by computer simulation.

There are some problems in VLSI layout that can be dealt with by combinatorial optimization; many among them are NP-complete problems. Using traditional algorithms. such as itcrative impproving,branch and bound, divide and conquer, we obtain too frequently only local minima. In this paper, we use a stochastic parallel algorithm to obtain optimal coveragein logic minimization which is a NP-complete problem in VLS layout. In implementing such an algorithm, careful consideration is given by the authors to the following...

There are some problems in VLSI layout that can be dealt with by combinatorial optimization; many among them are NP-complete problems. Using traditional algorithms. such as itcrative impproving,branch and bound, divide and conquer, we obtain too frequently only local minima. In this paper, we use a stochastic parallel algorithm to obtain optimal coveragein logic minimization which is a NP-complete problem in VLS layout. In implementing such an algorithm, careful consideration is given by the authors to the following three aspects: construction of objective functions. rule of choice of parameters, and choice of annealing schedule. We have used the algorithm to solve other VLSI layout problems,such as: circuit partition, assignment problem in gate array layout,gate sequencing in gate matrix layout and placement of blocks in building block layout. The results we obtain show that the algorithm bas low time complexity and low spare complexity with high parallelism and also that it is a good way to solve YLSI layout problems. It will be more effective if running on parallel computers.