压力和流量关联的神经网络解耦控制

Neural Network Decoupling Control of Coupling of Pressure and Flow

传统的解耦控制综合方法,对耦合对象的数学模型依赖性较大,因此限制了它在工业领域内的应用。本文以压力-流量串联系统为研究对象,将神经网络用于耦合对象的解耦控制之中,其思路是通过神经网络对系统的I/O数据的学习,不断修正自身的权值,达到解耦控制的目的。实验结果表明,神经网络可以使对象不确定的耦合系统实现解耦控制,效果令人满意。

Traditional decoupling controlling comprehensive method is quit dependent on mathematical model of coupling object, so its application is limited in industry control field.With the powerflow system,neural network is applied to decoupling control of coupling object,through constant adjusting selfweight value during the study from I/O data of the system,in order to get the aim of decoupling control. Simulating results prove that neural network decoupling control is satisfactory in the system with uncertain parameters.