A hybrid constriction coefficient-based particle swarm optimization and gravitational search algorithm for training multi-layer perceptron

PurposeIn this paper, a newly proposed hybridization algorithm namely constriction coefficient-based particle swarm optimization and gravitational search algorithm (CPSOGSA) has been employed for training MLP to overcome sensitivity to initialization, premature convergence, and stagnation in local optima problems of MLP.Design/methodology/approachIn this study, the exploration of the search space is carried out by gravitational search algorithm (GSA) and optimization of candidate solutions, i.e. exploitation is performed by particle swarm optimization (PSO). For training the multi-layer perceptron (MLP), CPSOGSA uses sigmoid fitness function for finding the proper combination of connection weights and neural biases to minimize the error. Secondly, a matrix encoding strategy is utilized for providing one to one correspondence between weights and biases of MLP and agents of CPSOGSA.FindingsThe experimental findings convey that CPSOGSA is a better MLP trainer as compared to other stochastic algorithms because it provides superior results in terms of resolving stagnation in local optima and convergence speed problems. Besides, it gives the best results for breast cancer, heart, sine function and sigmoid function datasets as compared to other participating algorithms. Moreover, CPSOGSA also provides very competitive results for other datasets.Originality/valueThe CPSOGSA performed effectively in overcoming stagnation in local optima problem and increasing the overall convergence speed of MLP. Basically, CPSOGSA is a hybrid optimization algorithm which has powerful characteristics of global exploration capability and high local exploitation power. In the research literature, a little work is available where CPSO and GSA have been utilized for training MLP. The only related research paper was given by Mirjalili et al., in 2012. They have used standard PSO and GSA for training simple FNNs. However, the work employed only three datasets and used the MSE performance metric for evaluating the efficiency of the algorithms. In this paper, eight different standard datasets and five performance metrics have been utilized for investigating the efficiency of CPSOGSA in training MLPs. In addition, a non-parametric pair-wise statistical test namely the Wilcoxon rank-sum test has been carried out at a 5% significance level to statistically validate the simulation results. Besides, eight state-of-the-art meta-heuristic algorithms were employed for comparative analysis of the experimental results to further raise the authenticity of the experimental setup.

Combining inertia and constriction technique in the PSO applied to fault identification in a hydraulic system

A novel methodology for simultaneous parametric fault isolation and mode switching identification in the spool motion of a Directional Control Valve (DCV), under multi fault assumption, has been reported in this paper. The shape of the profile traversed by the DCV spool was assumed to be trapezoidal in both healthy and faulty condition, but the slope of trapezoidal may change due to fault. Under this assumption, the real valued fault parameter and binary mode switching were identified by real valued Particle Swarm Optimization (PSO) alone instead of a combined real and binary valued PSO (Hybrid PSO). A novel PSO algorithm by combining the concepts of varying inertia weight (both increasing and decreasing trend) and constriction factor has been proposed in the article to achieve more accurate identification. Its validity was examined using an existing heuristic formula and by conducting several tests on a benchmark function used for fault identification. Superior improvement was observed in the identification with increasing inertia weight than that of widely used decreasing inertia weight, when combined with the constriction coefficient. A high pressure hydraulic circuit used in dumper and several other lifting machines was used as a simulation example.

PID Controller Design of Nonlinear System using a New Modified Particle Swarm Optimization with Time-Varying Constriction Coefficient

The proportional integral derivative (PID) controllers have been widely used in most process control systems for a long time. However, it is a very important problem how to choose PID parameters, because these parameters give a great influence on the control performance. Especially, it is difficult to tune these parameters for nonlinear systems. In this paper, a new modified particle swarm optimization (PSO) is presented to search for optimal PID parameters for such system. The proposed algorithm is to modify constriction coefficient which is nonlinearly decreased time-varying for improving the final accuracy and the convergence speed of PSO. To validate the control performance of the proposed method, a typical nonlinear system control, a continuous stirred tank reactor (CSTR) process, is illustrated. The results testify that a new modified PSO algorithm can perform well in the nonlinear PID control system design in term of lesser overshoot, rise-time, settling-time, IAE and ISE.Keywords: PID controller, Particle Swarm Optimization (PSO),constriction factor, nonlinear system.