A novel hybrid spiral dynamics bacterial chemotaxis algorithm for global optimization with application to controller design

2015

Applied Soft Computing

This paper presents hybrid spiral-dynamic bacteria-chemotaxis algorithms for global optimisation and their application to control of a flexible manipulator system. Spiral dynamic algorithm (SDA) has faster convergence speed and good exploitation strategy. However, the incorporation of constant radius and angular displacement in its spiral model causes the exploration strategy to be less effective hence resulting in low accurate solution. Bacteria chemotaxis on the other hand, is the most prominent strategy in bacterial foraging algorithm. However, the incorporation of a constant step-size for the bacteria movement affects the algorithm performance. Defining a large step-size results in faster convergence speed but produces low accuracy while defining a small step-size gives high accuracy but produces slower convergence speed. The hybrid algorithms proposed in this paper synergise SDA and bacteria chemotaxis and thus introduce more effective exploration strategy leading to higher accuracy, faster convergence speed and low computation time. The proposed algorithms are tested with several benchmark functions and statistically analysed via nonparametric Friedman and Wilcoxon signed rank tests as well as parametric t-test in comparison to their predecessor algorithms. Moreover, they are used to optimise hybrid Proportional-Derivative-like fuzzy-logic controller for position tracking of a flexible manipulator system. The results show that the proposed algorithms significantly improve both convergence speed as well as fitness accuracy and result in better system response in controlling the flexible manipulator.

Section: Referencesmentioning

confidence: 99%

Novel metaheuristic hybrid spiral-dynamic bacteria-chemotaxis algorithms for global optimisation

2015

Applied Soft Computing

“…Some of these have successfully employed the improved SDA to solve real world problems. Nasir et al, (2012a) have proposed hybrid spiral-dynamics bacteria-chemotaxis (HSDBC) in which a chemotaxis scheme of bacteria is incorporated into SDA to enhance the exploration strategy. The bacteria motion is designed such that they tumble and swim in a spiral form with the same or different value of spiral radius or angular displacement.…”

Section: Spiral Dynamics Algorithmmentioning

confidence: 99%

A novel hybrid bacteria-chemotaxis spiral-dynamic algorithm with application to modelling of flexible systems

Engineering Applications of Artificial Intelligence

2014

This paper presents a novel hybrid optimisation algorithm namely HBCSD, which synergises a bacterial foraging algorithm (BFA) and spiral dynamics algorithm (SDA). The main objective of this strategy is to develop an algorithm that is capable to reach a global optimum point at the end of the final solution with a faster convergence speed compared to its predecessor algorithms. The BFA is incorporated into the algorithm to act as a global search or exploration phase. The solutions from the exploration phase then feed into SDA, which acts as a local search or exploitation phase. The proposed algorithm is used in dynamic modelling of two types of flexible systems, namely a flexible robot manipulator and a twin rotor system. The results obtained show that the proposed algorithm outperforms its predecessor algorithms in terms of fitness accuracy, convergence speed, and timedomain and frequency-domain dynamic characterisation of the two flexible systems.

“…The algorithm was applied to optimisation of linear dynamic model of a flexible system. In another work, the authors proposed the same hybrid scheme but instead of using random swim, the bacteria swum in a spiral manner [8]. The proposed hybrid scheme was employed to optimize a proportional-derivative (PO) controller for a flexible manipulator system.…”

Section: Introductionmentioning

confidence: 98%

An improved spiral dynamic algorithm and its application to fuzzy modelling of a twin rotor system

2014 World Symposium on Computer Applications &Amp; Research (WSCAR)

Omar

et al. 2014

Self Cite

This paper presents the development of an improved spiral dynamic optimisation algorithm with application to fuzzy logic dynamic modelling of a twin rotor system. Spiral motion and dynamic step size generated by a spiral model produce unique exploration and exploitation strategies of a spiral dynamic algorithm. However, the algorithm is subj ect to settling into local optima at the end of the search process due to insufficient exploration throughout the search area. An elimination and dispersal strategy of a bacterial foraging algorithm is adopted to solve the problem. Moreover, the application of acquiring nonlinear dynamic model of a helicopter model prototype in hovering mode is presented and the results show the effectiveness of the proposed algorithm to solve real world problems. The result of the modelling work is presented in both time-domain and frequency-domain. It shows that the fuzzy model optimized by the proposed algorithm is better and adequate to represent the characteristic behaviour of the helicopter model prototype system.