Decoupled Control Strategy for Quadruple Tank Process

In a multivariable process mathematical models of the physical system have gained importance due to the complex interactions in the system. All the parameters used in a model cannot be determined accurately. However, such processes are very often available in power plants, chemical plant and many other fields. The major problem in a multivariable process is that loop interaction can arise and cause difficulties in feedback control design. This problem can be solved using centralized (like LQR, multivariable predictive control, IMC, fuzzy logic controller…) or decentralized controllers. The decentralized control is widely used in industrial plant. One of the decentralized techniques is decoupled method, which used to eliminate the interaction of diagonal loops and can be minimized an MIMO into SISO system. The basic and well known method for MIMO system is a two input two output (TITO) system.
In this paper, a new decoupler method with auto tuning of PID controller for quadruple tank process is proposed. It consist of four inter connected water tanks and two pumps as shown in figure 1. The parameters of PID controller are designed based on stability and speed of response, which are defined in terms of the damping factor and settling times. When damping factor and settling time of the controlled system has been selected the equivalent initial P, I, D values is obtained.