摘要
针对电子装备故障数据小样本、非线性的特点,在相空间重构处理原始时间序列数据的基础上,基于k折交叉验证和果蝇算法优化灰色神经网络模型参数,从而提出一种基于果蝇算法和灰色神经网络的故障预测方法,并以某型雷达高压电源监测数据仿真结果为例验证其模型性能;实验结果表明,相比已有方法,该方法在全局优化、收敛速度、预测精度方面都具有一定优势。
In view of the small sample and nonlinear characteristics of fault data in electronic equipment,the fault prognostic method of FOA and GNN is presented Fruit fly optimization algorithm （FOA） and k-fold cross validation are applied to optimize the model parameters of Grey Neural Network （GNN） with the original time sequence data are reconstructed in the phase space.And the performance of the proposed model is validated by the simulation results of high voltage power supply.Compared to other methods,the proposed model has better global optimization,convergence speed and prediction accuracy.
出处
《计算机测量与控制》
2015年第9期3081-3084,共4页
Computer Measurement &Control
基金
国家自然科学基金(No.61271153)
关键词
果蝇算法
灰色神经网络
k折交叉验证
电子装备
故障预测
fruit fly optimization algorithm （FOA）
grey neural network （GNN）
k-fold cross validation
electronic equipment
fault prognostic
