There is strong intercoupling relation between condenser water level and deaerator water level in marine steam power plant. In order to get satisfactory control effect, it is essential to take corresponding decoupling measures. PID neural network not only has the advantages of PID, but also has the abilities of learning, remembering and nonlinear approximation. In this paper, we propose a deaerator water level and condenser water level decoupling control strategy based on PID neural network, which integrates PID and neural network by establishing proportional neuron, integral neuron and derivative neuron corresponding to proportional, integral and derivative, respectively. We also propose a method of choosing initial weights and learning coefficient from experience of PID to enhance the convergence performance of PID neural network. The simulation results show that the PID neural network decoupling control strategy can meet the requirements of multivariable system decoupling control. It is more effective in condenser water level and deaerator water level decoupling than PID control strategy.
[1]
Zheng Yong.
The Control System of Multi-variable PID Neural Network and Its Application in Deaerator Water Level Control
,
2007
.
[2]
Xiaodong Liu,et al.
The PID Controller Based on the Artificial Neural Network and the Differential Evolution Algorithm
,
2012,
J. Comput..
[4]
Jing Wang,et al.
Improved Artificial Fish Algorithm for Parameters Optimization of PID Neural Network
,
2013,
J. Networks.
[5]
Shu Huailin.
ANALYSIS OF PID NEURAL NETWORK MULTIVARIABLE CONTROL SYSTEMS
,
1999
.
[6]
C. Crowe,et al.
Engineering fluid mechanics
,
1975
.
[7]
Huai Lin Shu,et al.
Study on Multivariable System Based on PID Neural Network Control
,
2012
.