摘要
电力变压器局部放电源超声定位已经演化为一个多学科的研究课题,具有明确的应用背景和重要的理论研究价值。研究了油中局部放电时超声波传播特性,建立了局部放电源超声定位的信号参数估计模型,将超声定位问题转化为一个带约束的非线性连续函数优化问题,在最优化框架内,应用了国际上最近提出的粒子群优化(PSO)算法进行求解。算例表明,文中算法具有计算效率高、通用性强、收敛速度快的优点,能有效防止结果陷入局部最优,计算结果明显优于最小二乘法。
The partial discharge source localization by ultrasonic measuring in a power transformer has been evolved into a multidisciplinary subject with a definite application background and important theoretical research value. In this paper, the signal parameter estimation model for ultrasonic localization has been developed by referring to the propagation properties of the ultrasonic wave. Thus, the localization problem is transformed into a nonlinearly continuous function optimization problem. Within the framework of optimization and inspired by certain mechanisms in sociology, psychology and ecology, the problem is solved with the recently proposed particle swarm optimization （PSO） algorithm. The results of calculating examples show that the solution of the algorithm proposed possesses high calculating efficiency and converging speed. PSO can effectively prevent the results from becoming locally optimized, which is evidently better than the least-square algorithm.
出处
《电力系统自动化》
EI
CSCD
北大核心
2005年第18期66-69,共4页
Automation of Electric Power Systems
基金
国家自然科学基金资助项目(50477040)。
关键词
电力变压器
超声定位
局部放电
超声波传感器
粒子群优化算法
power transformer
ultrasonic localization
partial discharge
ultrasonic sensor
particle swarm optimization algorithm
