Abstract: Abstract: A modification to Linearly Decreasing Weight strategy in standard Particle Swarm Optimization is taken by introducing descending index and iterative threshold. The inertia weight varies non-linearily with the changing of currently iterative order, exponent descending rate and iterative threshold .The new method is tested with three representative benchmarks and a compare is made with the standard Particle Swarm Optimization as well as other advanced Particle Swarm Optimization.It is demonstrated that there are evident superiorities in computational precision, searching speed and steady convergence.

摘要： 摘要：引入递减指数和迭代阈值对基本粒子群算法中线性递减权策略进行了改进，在优化迭代过程中，惯性权重随当前迭代次数、指数递减率和迭代阈值非线性变化。对三种具有代表性的测试函数进行了仿真实验，并与基本粒子群算法以及其他改进的粒子群算法进行了比较，结果表明，文中所提的改进粒子群算法在搜优精度、收敛速度以及稳定性等方面有明显优势。