电光与控制, 2020, 27 (11): 75, 网络出版: 2020-12-25  

基于改进粒子群的高斯过程故障预测模型建立方法

Establishing Gaussian Process Fault Prediction Model Based on Improved PSO
作者单位
海军航空大学, 山东 烟台 264001
摘要
针对高斯过程超参数使用共轭梯度法求解导致的依赖初始值以及容易陷入局部最优解的问题, 提出了一种基于参数非线性动态调整策略的改进粒子群算法, 并应用到超参数求解中。首先, 提出参数动态调整策略, 针对不同的搜索阶段采用不同的参数; 然后,根据免疫思想的浓度调节机制, 提高算法全局搜索能力。仿真实验结果表明,所提算法在最优化求解中能够快速收敛到函数最优值, 同时该算法在求解高斯过程超参数中具有有效性和优越性, 可以为后期预测模型的建立提供精度保障。
Abstract
Aiming at the problems of relying on initial value and being easy to fall into local optimal solution when using conjugate gradient method to solve Gaussian process hyperparameters,this paper proposes an improved Particle Swarm Optimization (PSO) algorithm based on parametric nonlinear dynamic adjustment strategy and applies it for hyperparameter solving.Firstly,the parameter dynamic adjustment strategy is proposed,and different parameters are adopted for different search stages.Then,according to the concentration adjustment mechanism of immune thought,the global search ability of the algorithm is improved.The simulation results show that:1) The proposed algorithm can quickly converge to the optimal value of the function in the optimal solution;and 2) At the same time,the algorithm is effective for solving the Gaussian process hyperparameters,which can guarantee the accuracy for the establishment of later prediction model.

吕佳朋, 史贤俊, 王康. 基于改进粒子群的高斯过程故障预测模型建立方法[J]. 电光与控制, 2020, 27(11): 75. LYU Jiapeng, SHI Xianjun, WANG Kang. Establishing Gaussian Process Fault Prediction Model Based on Improved PSO[J]. Electronics Optics & Control, 2020, 27(11): 75.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!