Application of Ant Colony Algorithm in Multi-objective Optimization Problems

Abstract

In actual application and scientific research, multi-objective optimization is an extremely important research subject. In reality, many issues are related to the simultaneous optimization under multi-objective conditions. The research subject of multi-objective IntroductionOriginated from the design, planing scale, project adjustment and other essential decision issues of many complex systems in real life, multi-objective optimization has always been one of the important subjects of engineering practice and scientific research. In the aspects of computing science, decision science and operation science, there once appeared much certainty, randomization methods specialized for multi-objective optimization [1]. In recent year, along with the improvement in the calculating speed and capability of calculation devices, the application of intelligent evolutionary algorithms in multi-objective optimization, as with genetic algorithm, genetic planing and genetic program designing, etc., has gained wide confirmation, which is mainly because these evolutionary algorithms process intelligence features of self-adaptivity, self-directed learning and self-organizing [2].In terms of multi-objective optimization, usually there are conflicts and restrains among different targets of the optimization problem. Therefore, in order to maintain balance among these targets, the research aim of the algorithm is to try best to locate the optimal set near the actual Pareto front. Considering the followed decision step, the solutions in the solution set should be distributed as evenly as possible to increase the diversity of the possible solution. Considering the effect of actual application, the time period of searching the Pareto solution set should be as short as possible [3]. Most of the current studies adopt genetic algorithm to explore multi-objective optimization, and the amount of studies adopting emerging intelligence algorithm, such as ant colony algorithm, to explore multi-objective optimization is quite limited. With characteristics of implicit parallelism and intelligence, ant colony algorithm is quite suitable for optimization. It can be seen from the current research results that ant colony algorithm has good performance. It is proved by practice that the application of ant colony algorithm in solving single-objective problems is very successful [4]. However, there are still a lot of problems to solve in the application of ant colony algorithm in multi-objective optimization. Fields worthy of exploration include how to select the initial ant colony, how to construct Pareto optimal solution set, how to set the parameters of any colony algorithm, how to conduct simulation experiment and the verification of related theories, etc [5].This paper first explained the basic principles of multi-objective problems and ant colony algorithm in detail, based on which, it provided the complete procedure of multi-objective ant colony optimization, and with the simulation, test and analysis of the standard test function, and

Image Analysis using Color Co-occurrence Matrix Textural Features for Predicting Nitrogen Content in Spinach

“…A research conducted by Zhou and Hu [45] also showed superiority of ACO to solve image clustering optimization problem in which ACO has fastest optimization speed with high optimization quality. Based on the research to solve MOO problem by Li and Tian [46], it showed that ACO had better performance compared to evolutionary algorithms such as GA and PSO. ACO can converge the fitness function of MOO more quickly and accurately, and can also maintain the distributivity of the better solution.…”

Bit Error Rate (BER) QoS Attribute in Solving Wireless Pricing Scheme on Single Link Multi Service Network

“…Recent research is already discussed by [11] that focused on pricing scheme on wireless of multi class QoS network where by applying some classes in single link network with the various QoS change and connection change. Their result can prove that by increasing connection cost and QoS change along with the cost change, can benefit providers by applying bit error rate (BER) QoS attribute.…”

“…Problem concerning with the wireless network is not only focus on the pricing schemes but many aspect can be reviewed and discussed. Scheduling and routing are one of the probles occurring in optimizing the wireless network [9] or using heuristics method described in [10].Recent research is already discussed by [11] that focused on pricing scheme on wireless of multi class QoS network where by applying some classes in single link network with the various QoS change and connection change. Their result can prove that by increasing connection cost and QoS change along with the cost change, can benefit providers by applying bit error rate (BER) QoS attribute.…”