摘要
提出一种基于拟牛顿信赖域法的电力系统非线性状态估计法。用拟牛顿法构造海森矩阵,虽然只利用了目标函数的一阶导数信息和目标函数值信息,但由于保证了正定条件和拟牛顿条件,比解析求得海森矩阵更高效稳定。用信赖域法代替原来的线搜索方法求解下降方向和步长,减少了算法的计算时间。通过对多个节点系统的仿真测试,验证了该算法的有效性。
A kind of nonlinear state estimation method based on quasi-Newton trust region method is proposed.Quasi-Newton method is used to construct Hessian matrix.Although it only uses first-order derivate information and numerical information of the target function,but due to ensuring positive definition and quasi-Newton condition,the Hessian matrix acquired by using quasi-Newton method is more highly stable than that acquired by using analytical method.It is able to reduce calculating time of the algorithm by using trust region method instead of line search method to solve descent direction and step size.Simulating testing on multiple node systems verifies effectiveness of this algorithm.
出处
《广东电力》
2016年第2期70-75,共6页
Guangdong Electric Power
关键词
电力系统
状态估计
非线性
拟牛顿法
信赖域法
power system
state estimation
nonlinear
quasi-Newton
trust region method
