Mohammad Mohammadi Najafabadi; Habibeh Nazif; Fahime Soltanian
Abstract
This paper is motivated by high dose rate brachytherapy treatment planning problems which involve the specification of the movement schedule of a radiation source so that the target volumes are adequately covered with sufficient doses and organs at risk are not radiated beyond the clinical acceptance ...
Read More
This paper is motivated by high dose rate brachytherapy treatment planning problems which involve the specification of the movement schedule of a radiation source so that the target volumes are adequately covered with sufficient doses and organs at risk are not radiated beyond the clinical acceptance threshold. It utilizes four powerful multi-objective evolutionary algorithms (MOEA), which create a set of equally-weighted Pareto optimal solutions instead of only one and produce better results compared to other optimization methods. These algorithms include non-dominated sorting genetic algorithms, Pareto envelope-based selection algorithm, non-dominated ranking genetic algorithm, and strength Pareto evolutionary algorithm. The results indicate that the last algorithm uses the dependency between decision variables to solve them efficiently and is the best type of MOEA both in terms of convergence criteria and solution diversity maintenance for the brachytherapy problems.
Ali Nemati; Ali Alizadeh; Fahime Soltanian
Abstract
This paper presents a numerical solution of a class of fractional optimal control problems (FOCPs) in a bounded domain having a noise function by the spectral Ritz method. The Bernstein polynomials with the fractional operational matrix are applied to approximate the unknown functions. ...
Read More
This paper presents a numerical solution of a class of fractional optimal control problems (FOCPs) in a bounded domain having a noise function by the spectral Ritz method. The Bernstein polynomials with the fractional operational matrix are applied to approximate the unknown functions. By substituting these estimated functions into the cost functional, an unconstrained nonlinear optimization problem is achieved. In order to solve this optimization problem, the Matlab software and its optimization toolbox are used. In the considered FOCP, the performance index is expressed as a function of both state and control functions. The method is robust enough because of its computational consistency in the presence of the noise function. Moreover, the proposed scheme has a good pliability satisfying the given initial and boundary conditions. At last, some test problems are investigated to confirm the efficiency and applicability of the new method.
Control and Optimization
Ghasem Ahmadi; Mohammad Teshnehlab; Fahimeh Soltanian
Abstract
o enhance the performances of rough-neural networks (R-NNs) in the system identification, on the base of emotional learning, a new stable learning algorithm is developed for them. This algorithm facilitates the error convergence by increasing the memory depth of R-NNs. ...
Read More
o enhance the performances of rough-neural networks (R-NNs) in the system identification, on the base of emotional learning, a new stable learning algorithm is developed for them. This algorithm facilitates the error convergence by increasing the memory depth of R-NNs. To this end, an emotional signal as a linear combination of identification error and its differences is used to achieve the learning laws. In addition, the error convergence and the boundedness of predictions and parameters of the model are proved. To illustrate the efficiency of proposed algorithm, some nonlinear systems including the cement rotary kiln are identified using this method and the results are compared with some other models.
