预测RNA二级结构离散粒子群优化算法

Discrete particle swarm optimization for RNA secondary structure prediction

根据RNA二级结构预测问题实质和基本粒子群优化算法特性,提出一种离散粒子群优化算法模型.定义该模型中一个可变集合搜索空间,设计了基于此空间粒子群速度与位置更新公式及运算规则.采用局部精英粒子优化策略解决了粒子群算法易陷入局部最优的问题.实验结果表明,该算法在收敛速度和精度上都具有较好的性能.

With the RNA secondary structure prediction and the characteristics of particle swarm optimization, a new model of discrete particle swarm optimization is presented. In the model, searching space of variable set is defined. The operation rules and the update formulae of particle＇s velocity and location are re-designed on the searching space. A local elitist particle strategy is introduced in order to avoid being trapped in local optima. The experimental results demonstrate the superiority of the proposed model over the other methods in terms of solution quality and convergence rates.