Control Of Continuous Stirred Tank Reactor Research Articles

Hybrid Control of Continuous Stirred Tank Reactor

This paper describes the design of a hybrid control for continuous stirred tank reactor (CSTR), with parameter uncertainty and system disturbance by intelligent techniques. CSTR is characterized by nonlinear and time-varying behaviour. Control of CSTR is a challenging problem. The proposed hybrid controller consists of a genetic algorithm (GA) based proportional integral derivative (PID) control and an internal model control (IMC). The IMC is constructed with a backpropagation (BPN) algorithm based neural emulator for the model and an Takagi Sugeno Fuzzy based inverse adaptive neuro-fuzzy inference system (ANFIS) as the controller. The IMC provides a very good tracking control and leads to control performance degradation in the case of regulatory problem which is solved using the GA-based PID control. GA-based PID implements the characteristics of GA’s global optimization to optimize the PID’s control parameters: Kp, Ki, Kd to provide best control effect. The proposed model provides an effective controllable range and gives robust control performance.

Artificial Neural Networks Based Modeling and Control of Continuous Stirred Tank Reactor

Continuous Stirred Tank Reactor (CSTR) is one of the common reactors in chemical plant. Problem statement: Developing a model incorporating the nonlinear dynamics of the system warrants lot of computation. An efficient control of the product concentration can be achieved only through accurate model. Approach: In this study, attempts were made to alleviate the above mentioned problem using &#34Artificial Intelligence&#34 (AI) techniques. One of the AI techniques namely Artificial Neural Networks (ANN) was used to model the CSTR incorporating its non-linear characteristics. Two nonlinear models based control strategies namely internal model control and direct inverse control were designed using the neural networks and applied to the control of isothermal CSTR. Results: The simulation results for the above control schemes with set point tracking were presented. Conclusion: Results indicated that neural networks can learn accurate models and give good nonlinear control when model equations are not known.

Control of CSTR using model predictive controller based on mixture distribution

The paper presents application of model-based predictive controller based on mixture distribution, i.e. a set of parallel models, to Continuous Stirred Tank Reactor (CSTR) process. This process is highly non-linear. Presented MPC is based on a mixture of linearized models in several operating points parameterised by their probabilities estimated on-line. Provided that the operating point of the process lies within the given set defined by convex envelope of the operating points presented MPC ensures “soft switching” between the individual models.

Temperature Control of Continuous Stirred-Tank Reactors by Manipulation of Fresh Feed

When reactor capacity is limited by heat removal, an often-recommended control structure is to run with maximum coolant flow and manipulate feed flow rate to control reactor temperature (TR ← F control). This control scheme has the potential to achieve the highest possible production rate; however, if the feed temperature is lower than the reactor temperature, the transfer function between temperature and feed flow rate contains a positive zero, which can degrade dynamic performance. This paper presents a quantitative study of this process, which is an example of the often encountered conflict between steady-state economics and dynamic controllability. The reactor is a jacket-cooled continuous stirred-tank reactor with an irreversible, exothermic, liquid-phase reaction A → B. The effects of feed temperature, design conversion, heat of reaction, heat-transfer coefficient, and design production rate are explored. Results indicate that the TR ← F control structure can only be used in reactors designed for fairly high conversions and for reactions with moderate heats of reaction. Control may be improved by operating with a feed temperature that is higher than the reactor temperature, despite the possible undesirable steady-state economic disadvantages of such a design.

Design of Fuzzy Control for Maintaining a CSTR in the Steep Section of Its Nonlinear Characteristic

The operation of a continuous stirred tank reactor (CSTR) in the steep domain of its nonlinear process characteristic is studied, using catalyzed hydrogen peroxide decomposition as a model process because of its transparent nonlinear dynamics. Unstable equilibrium states and corresponding limit cycles exist in the steep domain, the latter representing dynamically stable modes of operation where safer control is expected due to the lower educt concentration in the reactor. For CSTR control in the steep domain, a fuzzy approach is applied based on simulations and experimental understanding of the stable and unstable states of the reactor. Simulations are carried out for the set-point control of an unstable equilibrium state, and the tracking control of the corresponding stable limit cycle. The simulation results show that operation in the limit cycle requires much less control energy than that of the corresponding unstable equilibrium state in the case of disturbances, e. g., due to unavoidable modelling errors.

Linearizing Control of CSTR and Its Robust Stabilization by H.INF., Controller.

We aim to establish a robust CSTR control system with uncertain dynamics by combining linearizing control and H∞ control systems. The procedure presented in this study demonstrates practical usefulness for achieving a stable control system design for chemical plants such as a CSTR (Continuous Stirred Tank Reactor) which has not only strong nonlinearity, but also uncertain dynamics.

Control of continuous-stirred tank reactors: Stabilization with unknown reaction rates

In this work we study the stabilization of continuous stirred tank reactors with unknown reaction rates. We prove that, under certain general assumptions, the chemical reactor can be stabilized with high-gain feedback. Simulations show that high-gain feedback may lead to hard input saturation and poor closed-loop performance. A reaction rate uncertainty estimator is introduced which, when coupled with a nonlinear feedback, leads to global stability with acceptable control performance. We conclude that state observation is unnecessary for the stabilization of the class of chemical reactor considered here.

Remarks on Some Applications of Nonlinear Control Techniques to Chemical Processes

Abstract This paper is a collection of heterogeneous remarks on some typical nonlinear control problems of chemical plants. The time-scale control method used in [20] for dual composition control of distillation columns constitutes a simple way to avoid singularity of the decoupling matrix. For several continuous stirred tank reactors (CSTR) [17, 2], the linearizing outputs [13] corresponding to dynamic or static feedback linearization have a clear physical interpretation. The stability of the zero dynamics, associated with temperature control of CSTR [19] via input-output linearization techniques, is investigated: new necessary conditions, ensuring the global asymptotic stability, are given; they are directly expressed on the reaction scheme and kinetics. Saddle-node bifurcation control [7] of an exothermic CSTR is recalled. These remarks illustrate the crucial role played, for applications, by the qualitative analysis of the process dynamics.

A Feedforward/Feedback Control Synthesis for a Continuous Stirred Tank Reactor

For the purpose of synthesizing a feedforward/feedback control system of a continuous stirred tank reactor(CSTR), it is proposed first to use a linear time-varying time-delay model to represent the dynamics of CSTR which has a strong nonlinearity in the temperature and second to synthesize the control system for the CSTR by accomplishing a decoupling and disturbance rejection control in the feedforward loop and by supplementing a t-domain's multivariable Smith predictor in the feedback loop. The techniques employed in this study for the modeling of the nonlinear CSTR dynamics and for the synthesis of the feedforward/feedback control system have been proved to be useful for attaining a considerably favorable control of CSTR.