| 航空电子传感器系统Powell网络评估方法 |
| |
| 引用本文： | 杜昌平,周德云,田涛.航空电子传感器系统Powell网络评估方法[J].电光与控制,2006,13(6):6-8. |
| |
| 作者姓名： | 杜昌平 周德云 田涛 |
| |
| 作者单位： | 1. 西北工业大学,西安,710072
2. 成都飞机设计研究所,成都,610041 |
| |
| 摘 要： | 航空电子传感器系统是由多个子系统组成的复杂的系统,传统的传感器系统评估方法不适合评估航空电子传感器系统。为此,本文提出了一种基于Powell网络的航空电子传感器系统评估方法。该方法用两个独立的Powell网络对航空电子传感器输出信号进行一步预测,然后将两个预测值和系统输出信号当前实际量测值进行分析评估,并给出航空电子传感器系统的评估结果和系统可靠输出值。仿真实验和某型航空电子系统的实际应用表明:本文提出的传感器系统评估方法可行,并取得了理想的效果。
|
| 关 键 词： | Powell网络 系统评估 神经网络 遗传算法 航空电子传感器系统 |
| 文章编号： | 1671-637X（2006）06-0006-03 |
| 收稿时间： | 2005-09-07 |
| 修稿时间： | 2005-10-06 |
A Powell network based evaluation method for avionic sensor system |
| |
| DU Chang-ping,ZHOU De-yun,TIAN Tao.A Powell network based evaluation method for avionic sensor system[J].Electronics Optics & Control,2006,13(6):6-8. |
| |
| Authors: | DU Chang-ping ZHOU De-yun TIAN Tao |
| |
| Abstract: | Avionic sensor system is a complex system consisting of multiple subsystems,and the traditional evaluation methods for sensor system are not fit for it.We put forward a method for avionic sensor system evaluation based on Powell network.Two independent Powell networks are adopted for predicting the output signal of avionic sensors.The two predicted values are evaluated and analyzed together with the current actual value of the evaluated signal.The evaluation result of avionic sensor system and the reliable output of the system are presented.We also used genetic algorithm to improve Powell training method,the network training speed was increased greatly.Simulation and actual application showed that the method put forward here for sensor system evaluation is feasible and can obtain fine effect. |
| |
| Keywords: | Powell network system estimation neural network genetic algorithms avionic sensor system |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
