毕业设计(论文)：神经网络PID在锅炉蒸汽压力中的应用解读.doc

江苏科技大学 本 科 毕 业 设 计（论文） 学 院 专 业 学生姓名 班级学号 指导教师 二零壹贰年六月 江苏科技大学本科毕业论文 神经网络PID在锅里蒸汽压力中的应用 Research on the application in the boiler steam pressure based on BP-PID control 摘 要 船用锅炉主蒸汽压力调节对象含有大惯性、大滞后环节，而常规PID 控制方法不具备自适应能力，所以很难满足实际的控制要求。采用基于BP神经网络的PID控制方法，设计PID控制器的在线调整控制系统，改善系统的动态性能，无论在理论上还是在实践上都具有重要意义。 通过对传统PID控制原理和BP神经网络的学习，应用Matlab软件设计基于BP神经网络的PID控制器，利用神经网络的自学习能力来实现最佳组合的PID控制，方法简单，易于实现。 仿真研究证明，BP神经网络结合传统PID控制方法在锅炉蒸汽压力中的应用，取得了良好的控制效果。采用传统PID控制算法的响应曲线，振荡较大，过渡时间长，超调量较大；而采用BP神经网络控制算法的响应曲线，无振荡，过渡时间短，无超调，且最先得到稳定输出，其控制效果明显优于传统的PID控制算法。 关键词：锅炉蒸汽压力；BP神经网络；PID Abstract Marine used boiler steam’s pressure regulator object has large inertia and lag characteristic. It is difficult for the object to achieve the actual control requirements, as the conventional PID controler does not have the adaptive capacity. The use of PID control method based on BP neural network is of great significance to improve the dynamic performance of the system both in theory and practice, for the method adjusts the control system by the online design of PID controller’s parameters. According to the PID control principle and BP neural network learning method, the way of using the neural network’s self-learning ability to achieve the best control effect of the object is simple, easy to implement by the MATLAB software. The better control effect of the algorithm is confirmed by simulation process. On the one hand, using the traditional PID control algorithm, there is a larger oscillation, the transition for a long time and overshoot, as is seen from the response curve. On the other hand, the response curve of the BP neural network control algorithm is of no oscillation, and the transition time is short, and no overshoot.