Research And Application Of Artificial Plant Optimization Algorithm

Artificial plant optimization algorithm (APOA) is a recent proposedevolutionary computation in which the growing process of one tree is mappedinto the optimized problem. In APOA, three new operators: photosynthesisoperator, phototropism operator and apical dominance operator are designed tosimulate three important phenomenon. Each branch is one separate individual,photosynthesis operator is used to produce the energy, and phototropismoperator guides to the growth direction, as well as apical dominance operatormakes some adjustments for the defined directions. Simulation results show thestandard version is effective and efficient.In this paper, the mechanism of photosynthesis operator is analyzed, andan effective operator is designed, the main contributions are listed as follows:(1) In the standard version of APOA, the light responsive curve ofphotosynthesis operator is selected as rectangular hyperbolic model which is onegeneral one. To further investigate the influence of light responsive curve, sevenclassical models are chosen to compare, they are: rectangular hyperbolic model,non-rectangular hyperbolic model, updated rectangular hyperbolic model,parabola model, straight line model and two exponential curve models. In eachexperiment, the corresponding model is used to take as the light responsivecurve. Simulation results show the average performance of parabola model isbest when compared with other six models.(2) Different light responsive curve may result in different performance,therefore, a combination of some different light responsive curves may increasethe effectiveness of photosynthesis operator. In this paper, the whole evaluationperiod is divided into different parts, while in each part one light responsivecurve is selected with experimental design. With orthogonal experimental design,an optimal combination model is determined which consists parabola model,updated rectangular hyperbolic model and straight line model.(3) To testify the performance of modified photosynthesis operator withparabola model and one combined models, the coverage optimization problem of wireless sense networks is used to test. The coverage problem is how todetect the goal area with some sensors. Simulation results show it is effective.