节点文献
调速系统中的通用单神经元控制及其算法的研究
Research of a Versatile Single Neuron’s Control and Algorithm on DC and AC Drivers
【作者】 霍展羽;
【作者基本信息】 西安理工大学 , 电力电子与电力传动, 2004, 硕士
【摘要】 智能控制方法在调速系统的应用非常广泛,研究一种具有较强通用性的智能控制方法具有重要的现实意义。本文针对应用于调速系统的一种智能控制方法——单神经元PID控制器的基本原理、算法和各种改进方案进行了深入研究与分析。在此基础上,重点指出了普遍存在的权值恶化问题,分析了产生原因,设计了改进的单神经元控制器。这种改进的单神经元控制器把系统的状态和算法结构综合考虑,在不同的状态应用不同的控制算法,较好解决了权值恶化问题。与现有的单神经元控制器相比,具有通用性较强、算法简单、稳定性较好的优点。
【Abstract】 Intelligent control methods’ using is extremely widespread in the power drive systems and it has important practical significance to research and develop a versatile intelligent control method. A deeply study and summarize on principle, algorithms and developments of the single neuron PID (SN-PID) controller, which using in the power drive systems, are included in this paper. By the theoretical researching, the limitations of traditional SN-PID especially on the weight value’s deterioration are pointed out and the causations are analyzed. Based on the theory analysis, an improved SN controller is designed. This improved SN controller synthesizes the consideration of the system states and algorithm structure, using different control algorithm on the each system state. By this, it obviates the weight value’s deterioration. Comparing with the others improved SN-PID controller, it has the advantages of simplicity, stableness and versatility.
- 【网络出版投稿人】 西安理工大学 【网络出版年期】2004年 03期
- 【分类号】TM921.5
- 【被引频次】1
- 【下载频次】152