逆变式等离子切割电源控制策略  

Control Strategy of Inverted Power Supply for Plasma Cutting

作　　者：贾德利[1] 尤波[1] 

机构地区：[1]哈尔滨理工大学自动化学院,黑龙江哈尔滨150080

出　　处：《控制工程》2009年第2期184-187,共4页Control Engineering of China

基　　金：哈尔滨市优秀学科带头人研究专项基金资助项目(200RFXXG015);哈尔滨市青年科学研究基金资助项目(2008RFQXG065)

摘　　要：以逆变式等离子切割电源为研究对象,对其强非线性、时变、多变量耦合的特点及工艺要求,提出基于PID-NN算法的控制策略,从非线性受控对象建模、多工艺参数解耦、控制器加速及精度提高等方面入手,对控制器进行性能优化。通过仿真及切割实验,与传统PID控制器和单一的BP神经网络控制器进行比较,验证了该方案使系统在鲁棒性、响应速度等方面有较大的改善,切割质量得到了明显提高,证明了该控制器在处理离散事件及多变量的解耦问题的优势。The inverter plasma cutting power supply is studied. A control strategy based on PID-NN algorithm is proposed, according to the requirements of cutting technology and some characters of the system, including strong nonlinearity and time-variability. The controller performence is optimized in aspects of the modeling of nonlinear-controlled object, the decoupling of process parameters, and the improvement of controller accelerate and accuracy. The strategy is compared with the traditional PID controller and BP neural network by simulation and cutting experiment. The results show that the robustness and response character of the system is optimized, the cutting quality is improved, and the advantage of this controller in dealing with discrete event and decoupling parameters is obtained.

关 键 词：等离子切割 逆变 PID神经网络 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]