Showing papers by "Mojtaba Ahmadieh Khanesar published in 2013"

Discrete binary cat swarm optimization algorithm

Abstract: In this paper, we present a new algorithm binary discrete optimization method based on cat swarm optimization (CSO). BCSO is a binary version of CSO generated by observing the behaviors of cats. As in CSO, BCSO consists of two modes of operation: tracing mode and seeking mode. The BCSO presented in this paper is implemented on a number of benchmark optimization problems and zero-one knapsack problem. The obtained results are compared with a number of different optimization problems including genetic algorithm and different versions of binary discrete particle swarm optimization. It is shown that the proposed method greatly improves the results obtained by other binary discrete optimization problems.

Observer-based indirect model reference fuzzy control system with application to control of chaotic systems

Abstract: This paper proposes a novel, observer-based, indirect model reference fuzzy control approach for nonlinear systems, expressed in the form of a Takagi Sugeno (TS) fuzzy model. Based on this model, an adaptive observer based, indirect model reference fuzzy controller is developed to deal with external disturbances. In contrast to what is seen in the literature on adaptive observer based TS fuzzy control systems, the proposed method is robust in the existence of bounded external disturbances and it is capable of tracking a reference signal rather than just regulation. The proposed method is simulated on the control of Chua's circuit and it is shown that it is capable of controlling this chaotic system with high performance.

Sliding mode type-2 fuzzy control of robotic arm using ellipsoidal membership functions

Abstract: Several papers claim that the performance of the type-2 fuzzy logic systems is superior over their type-1 counterparts, especially under noisy conditions. In order to show the effectiveness of the noise reduction capabilities of the type-2 fuzzy logic systems, a novel type-2 fuzzy membership function, ellipsoidal membership function, has recently been proposed. The novel membership function has certain values on both ends of the support and the kernel, and some uncertain values on the other values of the support. The parameters responsible for the width of uncertainty are decoupled from the parameters responsible for the center and the support of the membership function. In this study, a sliding mode control theory based learning algorithm has been proposed to tune the consequent part parameters tuning of the ellipsoidal type-2 fuzzy membership functions. The applicability of the novel membership function with the proposed novel parameter update rules has been shown on the control of a 2DOF robotic arm. The simulation results show that the type-2 fuzzy neural networks working in parallel with conventional PD controllers have the ability of controlling the robotic arm with a high accuracy especially under noisy conditions.

Type-2 Fuzzy neural networks for sliding mode Fuzzy control of nonlinear dynamical systems with adaptive learning rate

Abstract: This paper proposes a novel training method for type-2 fuzzy neural networks (T2FNN). The proposed control method benefits from a sliding mode training method with adaptive learning rate. The proposed control structure is a feedback error learning structure so consists of a conventional controller in parallel with a T2FNN The conventional controller is responsible to stabilize the system. The stability of the proposed training method and the adaptive learning rate is proved using an appropriate Lyapunov function. The adaptive learning rate makes it possible to control the system without prior knowledge about the upper bound of the states of the system and their derivatives. The proposed approach is tested on the velocity control of an electro hydraulic servo. The proposed controller is compared with type-1 fuzzy neural networks; it is shown that in the presence of noise type-2 fuzzy system outperforms its type-1 counterpart.

Estimation of the parameters of wavelet neural networks using simultaneous use of genetic algorithm and recursive least square

Abstract: In this paper, a novel identification scheme based on wavelet neural network structure is proposed. The objective function for identification considered in this paper is the sum of squared error. In order to optimize this objective, the genetic algorithm (GA) which is a global optimization is used for the parameters which appear nonlinearly in the wavelet structure. Recursive least square algorithm is used for the parameters which appear linearly in the output of wavelet neural network because it is known to be an optimal estimator for these parameters. The proposed training algorithm is used to identify chaotic system and a highly nonlinear dynamical system. Simulation results show that the proposed method identifies input/output data with higher performance in terms of sum of squared error when it is compared to gradient descent method.