|
标题：基于ELMAN神经网络PID控制的BLDCM调速系统设计 |
|
作者：牛学锋 |
|
| 2015年第7期 访问次数：248次 |
|
摘要：针对无刷直流电机时变性、非线性、耦合性等特性，设计改进Elman神经网络PID控制算法。改进算法融合分层递阶思想和小生境PSO算法思想，联合优化ELMAN神经网络结构及初始化参数，解决传统控制算法收敛速度慢、易早熟、需人工设置网络结构及初始参数等问题。引入自适应灾变因子提高寻优精度。仿真结果表明,使用改进算法优化PID 控制器可使BLCDM调节时间和超调量大幅减小,响应速度加快,具备较好的动态性能和较强的鲁棒性。
关键词：无刷直流电机；Elman神经网络PID控制；调速系统
Abstract: A new improved PID controller parameters optimization algorithm was presented, aiming at the BLDCM characteristics of nonlinearity, timevarying volatility and strong coupling. Based on Hierarchical Algorithm and Niche PSO algorithm, the proposed algorithm optimize the parameters of the Elman neural network structure and initialization parameters together, which greatly improves convergence speed, avoid network falls into a local minimum and sets the neural network parameters manually. By importing adaptive catastrophe operations, the algorithms converge more accurate. The simulation results show that the improved method can make BLCDM reduce adjusting time and overshoots greatly, accelerate response speed and have better dynamic performance and robustness.
Key words: BLCDM; Elman neural network PID control; speedregulation |
|
| |
|
