Font Size: a A A

Research On Chaotic Ant Swarm Optimization Algorithm And Its Application

Posted on:2010-06-11Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y Y LiFull Text:PDF
GTID:1118360278965455Subject:Cryptography
Abstract/Summary:PDF Full Text Request
Optimization problems are widely encountered in the fields of engineering technology, scientific research and economic management. It is to look for a set of parameters in definite restriction with the aim of minimizing or maximizing the objective function. According to different kinds of problems, different methods have been presented, such as Newton's method, conjugate gradient method, Lagrange Multiplier Method, and so on. These methods can nicely find local extreme in different problems. However, with the development of science and technology, actual optimization problems become more and more complex. People have found that it is not easy to find a satisfying analytic solution when optimization problems represent the characteristics of complexity, constraint, nonlinearity, multi-minimum, and difficult to model. Therefore, it is necessary to find an optimization algorithm suiting for large-scale parallel Operation with smart features. Swarm intelligence is this kind algorithm which is inspired collective behavior of social insects or other animals to solve complex optimization problems.As a newly developed swarm intelligence paradigm, chaotic ant swarm optimization (CASO) is a very promising optimization tool, with many advantages in high-dimensional problems or tasks that lack prior knowledge. Its basic idea is originated from the chaotic and self-organizing behavior of the ants in nature. From 2006, the CASO has been applied successfully to solve some problems, such as parameter identification, data fitting, integer programming, and so on. However, the CASO algorithm has also some disadvantages in lower precision and longer computational time when solving complex optimization problems. Although some improved methods, such as augment the swarm scale and add the search bounds, can enhance the optimization performance to some extent, it can't overcome the algorithm's lower precision and reduce the computational time and computer memory completely. Thus, this article concentrates the research on the CASO and its applications. The main contributions given in this dissertation are as follows:1. Aiming at the lower solution precision of the CASO, this paper proposes a variation on the CASO, called the modified chaotic ant swarm optimization, or MCASO, employing restricting strategy and learning strategy to significantly improve the performance of the CASO. Application of the MCASO on five benchmark functions shows a marked improvement in performance over the CASO.2. Considering the premature convergence of the CASO, this paper presents an improved chaotic ant swarm optimization (ICASO) with three strategies, which are comprehensive learning strategy, search bound strategy and refinement search strategy. The first two strategies are employed to update ants' positions, which preserve the diversity of the swarm so that the ICASO discourages premature convergence. In addition, the refinement search strategy is adopted to increase the solution quality in the ICASO. Simulations show that the ICASO significantly enhances solution accuracy and convergence stability of the CASO. Basing on these test results, the ICASO is used to optimize PID parameters. Simulation shows that its optimization results are better than the CASO.3. In order to avoid the premature convergence of the CASO and improve its search efficiency, hybrid chaotic swarm optimization (HCASO) is proposed in this paper. The new algorithm introduces preselection operator and discrete recombination operator into the CASO; meanwhile it replaces the best position found by own and its neighbors' ants with the best position found by preselection operator and discrete recombination operator in evolution equation. Through testing five benchmark functions with large dimensionality, the experimental results show the new method enhances the solution accuracy and stability greatly, as well as reduces the computational time and computer memory significantly when compared to the CASO. In addition, we observe the results can become better with swarm size increasing from the sensitivity study to swarm size. And we gain some relations between problem dimensions and swam size according to scalability study. Finally, the HCASO is used to optimize the embedding strength in the digital watermark, simulations show the HCASO can implement the optimization embedding of watermark.4. This paper uses the CASO to solve equations. Based on distribution rules of algebra equations, the CASO was used for resolving complex equations and transcendental equations in an optimization way. Simulations show this method has properties of independence of initial iterative points, better adaptability and high precision. And it is an efficient tool in solving equations.5. A novel method of data fitting via the CASO is presented in this paper. Through the construction of a suitable function, the problem of data fitting can be viewed as that of parameter optimization, and then the CASO is used to search the parameter space so as to find the optimal estimations of the system parameters. To investigate the performance of the CASO, the CASO is compared with particle swarm optimization (PSO) on two test problems. Simulations indicate that the CASO achieves better performances.6. This paper uses the CASO to solve the problem of integer programming by embedding the search space Z~l into R~l and truncating the real values to the nearest integers. Two novel methods based on the CASO, rounding the real solutions after every step (CASO-F) and after the final step (CASO-S), are given and numerical experiences show that in most cases the CASO-F outperforms the CAS-S. Finally the CASO-F is compared with the PSO on seven widely used integer programming test problems. Results not only indicate that the CASO-F can efficiently deal with this problem, but also the CASO-F is better than the PSO in low dimension.
Keywords/Search Tags:swarm intelligence, chaotic ant swarm optimization, parameter identification, digital watermark, data fitting, integer programming
PDF Full Text Request
Related items