The neural network can be used to solve constrained optimization problems for multiple input and output variables. In the constrained system optimization, ordinary methods, such as linear and nonlinear programming and statistical regression, have encountered many difficulties. In contrast, the artificial neural network (ANN) has shown success in performing such tasks. ANN technology offers many opportunities in the performance optimization of fossil power plant systems. ANN can learn the performance characteristics of those systems from the regular monitoring or testing data. Plant performance tradeoffs can be predicted based on the ANN simulation. A PC-based computer code with a fast-learning algorithm application was developed to assist the system tuning. A combustion optimization example is presented to demonstrate the effectiveness of using this software to achieve the NO/sub x/ reduction and preserve the other performance parameters.
[1]
Robert Tibshirani,et al.
An Introduction to the Bootstrap
,
1994
.
[2]
Bernard Widrow,et al.
30 years of adaptive neural networks: perceptron, Madaline, and backpropagation
,
1990,
Proc. IEEE.
[3]
H. S. Hou,et al.
Neural network boiler optimization of efficiency, emission, and reliability with TVA Kingston Unit 3 low NOx optimization test results
,
1995
.
[4]
Susumu Sakakibara.
A Practice of Data Smoothing by B-spline Wavelets
,
1994
.
[5]
C. Chui.
Wavelets: A Tutorial in Theory and Applications
,
1992
.