| 遗传算法优化的神经网络频谱预测模型训练 |
| |
| 引用本文： | 杨 健,赵杭生,陈 曦.遗传算法优化的神经网络频谱预测模型训练[J].解放军理工大学学报,2016(6):505-511. |
| |
| 作者姓名： | 杨 健 赵杭生 陈 曦 |
| |
| 作者单位： | 1. 解放军理工大学 通信工程学院,江苏 南京 210007；2.中央军委装备发展部第63研究所,江苏 南京 210007； 3.南京理工大学 机械工程学院,江苏 南京 210094,1. 解放军理工大学 通信工程学院,江苏 南京 210007；2.中央军委装备发展部第63研究所,江苏 南京 210007； 3.南京理工大学 机械工程学院,江苏 南京 210094,1. 解放军理工大学 通信工程学院,江苏 南京 210007；2.中央军委装备发展部第63研究所,江苏 南京 210007； 3.南京理工大学 机械工程学院,江苏 南京 210094 |
| |
| 基金项目： | 国家自然科学基金资助项目(61471395,61301161,61471392)；江苏省自然科学基金资助项目(BK20141070)；国家863计划资助项目(2015AA7124068A) |
| |
| 摘 要： | 针对认知无线网络(CRN)中神经网络频谱预测模型原有的反向传播算法预测准确率不高的问题,提出一种遗传算法优化的神经网络频谱预测模型训练(GA-NN)算法。通过在选择、交叉、变异中加入随机性,使种群的个体收敛至包含全局最优解的集合内,再通过反向传播算法训练神经网络频谱预测模型(BPNN)快速搜索到全局最优解。仿真结果表明,GA-NN算法训练的神经网络频谱预测模型的预测准确率比BP-NN算法提高一倍以上,GA-NN算法在多种CRN中具有适用性。GA-NN算法提高了频谱预测模型的预测准确率,将促进频谱预测技术在CRN中的推广应用。
|
| 关 键 词： | 频谱预测 遗传算法 神经网络 局部最优解 认知无线网络 |
| 收稿时间： | 2016/4/29 0:00:00 |
| 修稿时间： | 2016/6/17 0:00:00 |
Neural network spectrum prediction model training of genetic algorithm optimization |
| |
| YANG Jian,ZHAO Hangsheng and CHEN Xi.Neural network spectrum prediction model training of genetic algorithm optimization[J].Journal of PLA University of Science and Technology(Natural Science Edition),2016(6):505-511. |
| |
| Authors: | YANG Jian ZHAO Hangsheng and CHEN Xi |
| |
| Institution: | 1. College of Communications Engineering, PLA Univ. of Sci. & Tech. , Nanjing 210007, China;2. The 63rd Research Institute of the Equipment Development Department of the Central Military Commission,Nanjing 210007,China; 3. School of Mechanical Engineering, Nanjing University of Science & Technology, Nanjing 210094, China,1. College of Communications Engineering, PLA Univ. of Sci. & Tech. , Nanjing 210007, China;2. The 63rd Research Institute of the Equipment Development Department of the Central Military Commission,Nanjing 210007,China; 3. School of Mechanical Engineering, Nanjing University of Science & Technology, Nanjing 210094, China) and 1. College of Communications Engineering, PLA Univ. of Sci. & Tech. , Nanjing 210007, China;2. The 63rd Research Institute of the Equipment Development Department of the Central Military Commission,Nanjing 210007,China; 3. School of Mechanical Engineering, Nanjing University of Science & Technology, Nanjing 210094, China |
| |
| Abstract: | In cognitive radio network(CRN), neural network based spectrum prediction model is trained by back propagation algorithm. To solve the low prediction accuracy problem, the genetic algorithm optimized neural network training for spectrum prediction(GA-NN) model was presented. Selection, crossover and mutation were applied to increasing the randomness of GA-NN training, which ensures the population converge to the set that contains the global optimal solution. Then, back propagation based neural network training for spectrum prediction(BP-NN) was conducted to search the global optimal solution efficiently. The simulation results show that the prediction accuracy is improved by 100% or more. The application of GA-NN in CRN was discussed for various scenarios. GA-NN training significantly improves the prediction accuracy for neural network spectrum prediction model, which will promote the application of spectrum prediction in CRN. |
| |
| Keywords: | spectrum prediction genetic algorithm neural network local optimal solution cognitive radio network |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《解放军理工大学学报》浏览原始摘要信息 |
| 点击此处可从《解放军理工大学学报》下载免费的PDF全文 |
|
