| 非线性系统的神经网络广义预测控制 |
| |
| 引用本文： | 杨志军,齐晓慧,单甘霖. 非线性系统的神经网络广义预测控制[J]. 军械工程学院学报, 2008, 20(3): 55-58 |
| |
| 作者姓名： | 杨志军 齐晓慧 单甘霖 |
| |
| 作者单位： | 军械工程学院光学与电子工程系,河北石家庄050003 |
| |
| 摘 要： | 研究了神经网络广义预测控制方法在非线性系统中的应用，基于BP网络构造神经网络预测器，利用非线性系统的开环输入输出数据离线训练神经网络，根据拟牛顿BFGS优化算法使得二次型性能指标函数达到最小，得到了最优的控制序列。同时给出了神经网络广义预测控制算法的步骤，讨论了提高系统鲁棒性的措施。仿真结果表明，这种神经网络预测控制算法具有响应速度快、控制效果好和跟踪精度高等特点。
|
| 关 键 词： | 非线性系统 神经网络 广义预测控制 鲁棒性 |
Neural Network Based Generalized Predictive Control of a Nonlinear System |
| |
| YANG Zhi-jun,QI Xiao-hui,SHAN Gan-lin. Neural Network Based Generalized Predictive Control of a Nonlinear System[J]. Journal of Ordnance Engineering College, 2008, 20(3): 55-58 |
| |
| Authors: | YANG Zhi-jun QI Xiao-hui SHAN Gan-lin |
| |
| Affiliation: | (Department of Optics and Electronics Engineering, Ordnance Engineering College, Shijiazhuang 050003, China) |
| |
| Abstract: | A neural network based generalized predictive controller is applied for a nonlinear system. First, a neural predictor is developed with a BP network. Second, the input data and output data of the nonlinear system are used for the network＇ s off-line training. Third, a Quasi-Newton method using the BFGS-algorithm is used to minimize the cost function. Finally, the control sequence is brained to control the system. This control algorithm steps are listed in detail and some measures to improve the robustness are discussed. The simulation results show this algorithm has the characteristics of rapid response, good control quality and high tracking accuracy. |
| |
| Keywords: | nonlinear system neural network generalized predictive control robustness |
| 本文献已被 维普 等数据库收录！ |
