Model-based Predictive Control Algorithm Research Articles

Model-based predictive control for biomass production in advanced life support

A model-based predictive control (MBPC) algorithm for crop production in a biomass production component of an Advanced Life Support System (ALSS) is introduced. The MBPC adjusts light intensity, air temperature, and atmospheric carbon dioxide concentration setpoints so that crop growth (total and yield biomass) follows pre-determined production schedules. Mathematical models developed for wheat, soybean, and white potato using multivariate polynomial regression (MPR) are used as the feed-forward models in the control. The control algorithm is evaluated for its ability to compensate for simulated environmental disturbances and sensitivity to modeling and measurement errors. A 24-hour time increment is utilized

A stabilizing model-based predictive control algorithm for nonlinear systems

Predictive control of nonlinear systems subject to state and input constraints is considered. Given an auxiliary linear control law, a good nonlinear receding-horizon controller should (i) be computationally feasible, (ii) enlarge the stability region of the auxiliary controller, and (iii) approximate the optimal nonlinear infinite-horizon controller in a neighbourhood of the equilibrium. The proposed scheme achieves these objectives by using a prediction horizon longer than the control one in the finite-horizon cost function. This means that optimization is carried out only with respect to the first few input moves whereas the state movement is predicted (and penalized) over a longer horizon where the remaining input moves are computed using the auxiliary linear control law. Closed-loop stability is ensured by means of a penalty on the terminal state which is a computable approximation of the infinite-horizon cost associated with the auxiliary controller. As an illustrative example, the predictive control of a highly nonlinear chemical reactor is discussed.

Model-based predictive control for Hammerstein?Wiener systems

In this paper a model-based predictive control (MPC) algorithm is presented for Hammerstein?Wiener systems. This type of system consists of a linear dynamic block preceded and followed by a static non-linear block. These systems appear useful in modelling several non-linear processes encountered in industry. Directly using such a model in a MPC algorithm will in general lead to a non-linear optimization problem due to the static non-linearities. This can be avoided by exploiting the structure of these models. In this paper the non-linearities are transformed into polytopic descriptions. This procedure enables one to use robust linear MPC techniques for controlling these systems. In such a way a convex optimization problem is retained. For the presented MPC algorithm, which is stated as an optimization problem subject to linear matrix inequalities, nominal closed loop stability is proven. In two examples it is shown that by means of transforming the non-linearities into polytopic descriptions, as done in the presented MPC algorithm, a better tuning of the input?output behaviour of the plant is obtained, compared to removing the static non-linearities from the control problem by an inversion, a technique often used for these systems.

A model-based algorithm for blood glucose control in type I diabetic patients.

A model-based predictive control algorithm is developed to maintain normoglycemia in the Type I diabetic patient using a closed-loop insulin infusion pump. Utilizing compartmental modeling techniques, a fundamental model of the diabetic patient is constructed. The resulting nineteenth-order nonlinear pharmacokinetic-pharmacodynamic representation is used in controller synthesis. Linear identification of an input-output model from noisy patient data is performed by filtering the impulse-response coefficients via projection onto the Laguerre basis. A linear model predictive controller is developed using the identified step response model. Controller performance for unmeasured disturbance rejection (50 g oral glucose tolerance test) is examined. Glucose setpoint tracking performance is improved by designing a second controller which substitutes a more detailed internal model including state-estimation and a Kalman filter for the input-output representation. The state-estimating controller maintains glucose within 15 mg/dl of the setpoint in the presence of measurement noise. Under noise-free conditions, the model-based predictive controller using state estimation outperforms an internal model controller from literature (49.4% reduction in undershoot and 45.7% reduction in settling time). These results demonstrate the potential use of predictive algorithms for blood glucose control in an insulin infusion pump.

Modifying the Prediction Equation for Nonlinear Model-Based Predictive Control

This paper compares the performance of three nonlinear model-based predictive control (NLMPC) algorithms on the basis of how they modify the standard linear MPC prediction equation to handle nonlinear systems. A simple static system and a simulated polymer reactor are used for illustrative purposes.

A neural net model-based multivariable long-range predictive control strategy applied in thermal power plant control

A constrained multivariable control strategy along with its application in more efficient thermal power plant control is presented in this paper. A neural network model-based nonlinear long-range predictive control algorithm is derived, which provides offset-free closed-loop behavior with a proper and consistent treatment of modeling errors and other disturbances. A multivariable controller is designed and implemented using this algorithm. The system constraints are taken into account by including them in the control algorithm using real-time optimization. By running a simulation of a 200 MW oil-fired drum-boiler thermal power plant over a load-profile along with suitable PRBS signals superimposed on controls, the operating data is generated. Neural network (NN) modeling techniques have been used for identifying global dynamic models (NNARX models) of the plant variables off-line from the data. To demonstrate the superiority of the strategy in a MIMO case, the controller has been used in the simulation to control main steam pressure and temperature, and reheat steam temperature during load-cycling and other severe plant operating conditions.

IDCOM - B multivariable model-predictive controller

In the last ten years Model Based Predictive Control (MBPC) has proved itself as one of the most successful advanced control tools. Its capability to handle processes with 'difficult' dynamics, has allowed improved control on critical loops with long time delays, constraints, feed-forwarding, over-rides, etc. Moreover, MBPC algorithms are able to perform multivariable control on processes characterised by strongly interactive loops, which have always been one of the toughest control problems. Due to the computational burden of such sophisticated algorithms, and the engineering skills required for a successful implementation, MBPC diffusion has been limited, in most cases, to 'money-making-loops', above all in the chemical and petrochemical fields, where even small increases in efficiency or yield produce major savings. The great progress in utilising the always increasing power of digital computers, and in developing and improving user-friendly software environments, provides opportunities for applying highly sophisticated control techniques more quickly, easily and cheaply. This paper introduces the new IDCOM-B™ controller; that is, the first multivariable model-predictive control package completely embedded in a DCS as a modular control function. Its unique capability to run inside a standard INFI90 control module, sharing with the base regulatory loops the communication channels, control database and operator interfaces, avoids the cost of an external computer and all the efforts required to interface it with a DCS architecture.

Nonlinear model-based predictive control of control nonaffine systems

This paper proposes a new nonlinear model-based predictive control (NLMPC) algorithm, designed to handle control nonaffine systems (nonlinear in the manipulated variable), which is based on a reinterpretation of the prediction equation as a Taylor series expansion. They key feature of this algorithm lies in the use of a process output prediction that accounts for changes in process dynamics as a function of the operating point as well as of the magnitude of the process input change. From this algorithm, a suboptimal NLMPC algorithm is derived to improve computational efficiency for real-time applications. The performance improvement of the new algorithm is illustrated by comparing it with a standard nonlinear model-based predictive control algorithm using three examples. The first two examples are simulations of simple static nonlinear systems, and the third example is a simulation of a semibatch acrylonitrile-butadiene (NBR) emulsion polymerization process.

Model-Based Predictive Control: Theory and Implementation Issues

Model-based predictive control algorithms have been widely applied in the chemical process industry. This paper explains in a tutorial manner, the control philosophy associated with such long range predictive or multi-step optimization strategies. One of the weak points in the development of this strategy has been the handling of constraints. For example, the control action implemented on a physical system has limitations due to saturation of actuators, i.e. a valve can only open between the range of 0% and 100%. In addition there are often quality constraints imposed on the process output and there may be constraints imposed on the incremental control action as well. The incorporation of these constraints into a model-based predictive control algorithm is illustrated by a geometric interpretation of the constrained and unconstrained solutions of this controller. Methods of handling such constraints, analytically or through external subroutine calls, are discussed together with new developments in this area with respect to the addition of a terminal matching condition in the control objective function. The paper also reviews current work concerned with control-relevant-identification techniques for long-range predictive control.