A novel approach to wide-range optimal reactor temperature control using diagonal recurrent neural networks (DRNNs) with an adaptive learning rate scheme is presented. The drawback of the usual feedforward neural network (FNN) is that it is a static mapping and requires a large number of neurons and takes a long training time. The usual fixed learning rate based on an empirical trial and error scheme is slow and does not guarantee convergence. The DRNN is for dynamic mapping and requires much fewer neurons and weights, and thus converges faster than FNN. A dynamic backpropagation algorithm coupled with an adaptive learning rate guarantees even faster convergence. The DRNN controller described includes both a neurocontroller and a neuroidentifier. A reference model which incorporates an optimal control law with improved reactor temperature response is used for training of the neurocontroller and neuroidentifier. Rapid convergence of this DRNN-based control system is demonstrated when used to improve reactor temperature performance. >

Improved nuclear reactor temperature control using diagonal recurrent neural networks

The authors present the design of a robust controller using the linear quadratic Gaussian with loop transfer recovery (LQG/LTR) for nuclear reactors with the objective of maintaining a desirable performance for reactor fuel temperature and the temperature of the coolant leaving the reactor for a wide range of reactor powers The results obtained are compared to those for an observer-based state feedback optimal reactor temperature controller Sensitivity analysis of the dominant closed-loop eigenvalues and nonlinear simulation are used to demonstrate and compare the performance and robustness of the two controllers The LQG/LTR approach is systematic, methodical, and easy to design and can give improved temperature performance over a wide range of reactor operation >

/pdf/lqg-ltr-robust-control-of-nuclear-reactors-with-improved-30rfssmrfp.pdf

LQG/LTR robust control of nuclear reactors with improved temperature performance

State feedback assisted classical (SFAC) control has been developed to improve the temperature response performance of nuclear reactors via modifying the embedded classical controller reference signal. This is done by means of an outermost state feedback controller. A linear quadratic Gaussian with loop transfer recovery (LQG/LTR) at the plant output seems a good candidate for the state feedback loop of SFAC structure, but it ends up in a closed loop system with tightly controlled power. To pay more attention to temperature responses, this paper presents the results of using LQG/LTR at the plant input in SFAC structure. We impose a minor change for considering, to some extent, the variation of system poles (and hence its speed) due to linearization of the nonlinear plant in equilibrium conditions other than the design power. The results are compared to an existing LQG controller with LTR at the plant output. Sensitivity of dominant closed loop poles and nonlinear simulations are used for demonstration and comparison.

Improved temperature control of a PWR nuclear reactor using an LQG/LTR based controller

Robust nonlinear model predictive control for a PWR nuclear power plant

A neural network controller for load following operation of nuclear reactors

A new control concept, state feedback assisted classical control, is described. The concept incorporates a classical output feedback control system as an easily understood inner control loop and ca...

State Feedback Assisted Classical Control: An Incremental Approach to Control Modernization of Existing and Future Nuclear Reactors and Power Plants

When gamma rays are used to charge particulate matter in stack gas, high collection efficiencies are obtained. These high efficiencies are achieved for all sizes of particulates, including those of...

M. A. Schultz

Papers

State Feedback Assisted Classical Control: An Incremental Approach to Control Modernization of Existing and Future Nuclear Reactors and Power Plants

Collection of Particulate Matter from Smokestacks Using Gamma-Ray Ionization