The controllers′ parameters cannot be adjusted directly by output error signals because of a secondary path. An improved simultaneous perturbation adaptive algorithm,which proves to be a stochastic gradient approximation method in terms of mathematical expectation,was presented.
Controller parameters cannot be updated directly with the output error signal due to the existence of secondary path,so a modified simultaneous perturbation algorithm,which is proved a stochastic gradient approximation method in the name of mathematic expectation,is presented to update controller weights without modeling secondary path.
Based on the principle of model free adaptive control(MFAC),simultaneous perturbation algorithm(SPA)is used to process the measured error signal of active noise control system directly,so it is unnecessary to model the secondary acoustic path.
The algorithm with simultaneous perturbation stochastic approximation method is the model-free control algorithm with stochastic approximation method and without modeling. It is superior to those industrial processes with non-linearity and great variations.
Bagging SVM and Boosting SVM cannot improve the classification result. In this dissertation, RBaggSVM and RBoostSVM algorithms manipulating both training sets and SVM model parameters are proposed. It is characteristic of randomization of SVM model parameters within specific scope for achievement of diversity component classifiers.
The intervention of Multiple Model Approach improve the whole system's rapidness, robustness and stability. The parameters of the multiple fixed model controller can re-initialize the adaptive controller which is based on the M2SPSA algorithm.
It is further found that the final maximum amplitude of an unstable ultra-long wave depends only on the structure of the initial state of the basic field, but not on the initial state of the disturbance.
fuzzy expected value model, fuzzy chance-constrained programming model, and fuzzy dependent-chance programming model, a simultaneous perturbation stochastic approximation algorithm is proposed by integrating neural network with fuzzy simulation.
This paper deals with the problem of the instability and development of the ultra-long waves in frictionless and adiabatic motion. It is found that the controlling factors on the instability are the vertical distribution of the static stability and the vertical shear of the basic current. The occurrence of the instability requires that the upper layer of the atmosphere possesses higher static stability than the lower layer. The existence of the stratosphere provides this requirement. Further the degree of i...
The local influence approach in linear regression models with two kinds of specific dispersion matrices is developed and discussed.An advantage of this approach is that it avoids the inappropriate case-deletion diagnostic in linear models witch discussed and it allows simultaneous perturbation on all responses. In addition,two simulation examples are given and discussed.
In this paper, we discuss the stability of super efficient points. In the sense of semicontinuity, the stability for super efficient points of a perturbation set with a perturbation cone is obtained in Banach space.