摘要
船舶蒸汽动力装置中除氧器压力和除氧器水位互相关联,具有很强的耦合性,传统的PID控制很难获得令人满意的控制效果,因此必须采取相应的解耦措施。PID型神经网络不仅具有传统PID的优点,还具有神经网络的自学习和逼近任意函数的能力。本研究建立了除氧器压力和水位的模型,并通过建立与比例、积分和微分相对应的神经元,将PID和神经网络整合在一起,提出一种PID型神经网络解耦控制方法。在所建立的除氧器压力和水位模型上对PID型神经网络解耦控制方法进行仿真。仿真结果表明,相对于单回路PID控制方法,该方法具有比单回路PID控制更好的解耦效果,可以将除氧器压力和水位的稳定时间分别缩短100s和60 s,并将二者的超调量分别减少0.6 KPa和0.005 m。
In marine steam power plants,the pressure and water level in deaerators are correlated and have a strong coupling property. As a result,it is very difficult for the traditional PID control to achieve satisfactory control effectiveness and it is mandatory to take corresponding decoupling measures. PID type neural networks not only have the merits of the traditional PID control but also have an ability of performing a self-learning and approaching to any function. A model for the pressure and water level in deaerators was established and through establishing a neuron corresponding to the proportional,integral and differential control,the PID control and the neural network were integrated and a PID type neural network decoupling control method was proposed. By making use of the model thus established,a simulation by using the PID type neural network decoupling control method was performed. It has been found that compared with the single loop PID control method,the method in question boasts a better decoupling result,the stabilization time durations of the pressure and water level in the deaerator can be shortened by 100 s and 60 s respectively and both overshoots can be reduced by 0. 6 KPa and 0. 005 m respectively.
出处
《热能动力工程》
CAS
CSCD
北大核心
2015年第6期926-931,976,共6页
Journal of Engineering for Thermal Energy and Power
关键词
蒸汽动力
除氧器压力
除氧器水位
PID型神经网络
多变量解耦控制
steam power,pressure in a deaerator,water level in a deaerator,PID type neural network,multi-variable decoupled control
