This paper concerns a study on the optimal control for nonlinear systems. An appropriate alternative in order to alleviate the nonlinearity of a system is the exact linearization approach. In this fashion, the nonlinear system has been linearized using input-output feedback linearization (IOFL). Then, by utilizing the well developed optimal control theory of linear systems, the compensated nonlinear system has been controlled. Thus, the structure of the objective function will be converted into a quadratic form which is qualitativly comparable with usual cost functions, and from operating viewpoint is more favorable. To qualify such a procedure, it has been applied to two minimum and nonminimum-phase chemical processes, and its performance is verified through computer simulations.

In this paper an objective orientation is presented for controlling multi-objective systems. The principles of this method is based on an emotional learning and temporal difference learning, and has a neuro-fuzzy structure. The proposal method can control the system in a way that objectives such as the present conditions, the system action in the part and the controlling aims are attained in the best way and least amount of time.The holistic structure of this paper is as follows: first, in the third unit object oriented control is studied. The fourth unit deals with emotional learning as a method for intelligent control. In the fifth unit the emotional control of temporal difference is defined and discussed. The sixth unit presents and discusses a new critic structure with temporal difference learning and then Some Multivariable systems, which are one of the most important controlling problems because of coupling, are defined. Then in the last unit the functional control structures used in this paper are studied and the experimental results are compared.

The design and analysis of Multivariable fuzzy logic feedback control systems are presented in this paper. A general design procedure is discussed and a multi-dimensional rule base in the form of the Fuzzy Relational Matrix (FRM) developed. The decision-making components of the Fuzzy Logic Controller(FLC) are not explicitly based on the conventional Fuzzy Associative Memory (FAM). The compositional rule of inference is employed to transform the set of fuzzy rules into the corresponding I;RM"s. A fuzzy connective OR (Maximum) is operated on the set of derived FRM"s to obtain a final fuzzy matrix which essentially constitutes the rule base of the FLC. This matrix establishes the relationship between a given controller input with every controller output. Hence, instead of using a forward changing technique for firing the applicable rule, an unknown output is evaluated by the composition of the related input with the FRM. Different MIMO control system configurations are investigated and direct interaction control as well as the cross-coupled fuzzy controllers implemented. Results of extensive simulation studies of two linked tanks are presented to illustrate the usefulness of the proposed approach.

In this paper, an innovative adaptive output feedback control scheme is proposed for general multi-input multi-output (MIMO) plants with unknown parameters in a regulation task; such that the outputs of the plant converge to zero as well as the control gains remain uniformly bounded. First an adaptive observer is designed to estimate the state variables and system parameters by using the inputs and outputs of the plant. Then a linear combination of the estimated states by adaptive control gains is used to design a robust adaptive controller. Some theorems are given to show the convergence of the modeling errors and the control gains. The proposed controller is used to control a two degree of freedom robot manipulator such that the robot moves from any initial configuration to zero position. Simulation results exhibit the effectiveness of the proposed scheme to control the robot manipulator with different initial conditions and parameter perturbations.

A QFT-type frequency-domain controller design technique is presented for multiple input-multiple ouput (MIMO) nonlinear systems with significant structured (parametric) uncertainty. Hard time-domain constraints are imposed on the outputs and control inputs in response to step disturbances. The control design method is based on replacing the nonlinear MIMO system with an equivalent diagonal MIMO linear uncertain plant using compact and convex sets of output functions. Then, a diagonal controller is designed for the equivalent linear system by transferring the time-domain constraints into the frequency-domain to obtain a series of allowable regions for the frequency responses of the nominal loop transfer functions. Next, a loop-shaping procedure is used to shape the nominal loop transfer function and hence the controller for each loop. Finally, Schauder"s fixed point theorem is utilized to show that the same controller will also work for the nonlinear system. The proposed method is illustrated by an example.

In this paper we use radial basis functions to solve Multivariable integral equations. We use collocation method for implementation. Numerical experiments show the accuracy of the method.

Interactions between input and output variables are a prevalent challenge in the design of multi-loop controllers for Multivariable processes, and they can be a major stumbling block to obtaining good overall performance of a multi loop control system. The deconstructed dynamic interaction analysis is proposed to solve this limitation by decomposing the multi loop control system into a series of n independent SISO systems, each with its own PID controller. The Multivariable decoupler and multi loop PID controller is applied to Two Tank Conical Interacting System (TTCIS). This TTCIS is chosen as benchmark problem used by many researchers. Firstly, the Mathematical modelling of TTCIS is derived using First principal model. The non-linear system is linearized using Jacobian matrix and decomposed into multiple SISO systems. The controller design for the process is then obtained, and an RGA matrix is constructed to minimise the interaction effects. To demonstrate the efficiency of the suggested strategy, simulation results using TTCIS are provided.

In this paper, a semi-active Multivariable adaptive controller is designed for a framed structure under a seismic disturbance using a filtered-x NLMS optimizer. Actuator and large power supply are not required, since just some valves and a battery-size power supply are sufficient for controlling the amount of hydraulic fluid flow through the by-pass on off-orifice channel in an energy dissipating system ,which consists of a piston attached to L-shaped wind bracing of the building and a cylinder attached to the upper floor. It is assumed that only a rough finite element model of the structure is available. In addition, in order to search for optimal non-classical viscous damp r coefficients, acting as calibration measures for the corresponding orifice size of the by-pass cannels of dissipation systems, the adaptive controller parameters are estimated by Normalized last-Mean-Squares strategy. The simulation results demonstrate that even though the controllability offered by this system does not equal that of an active counterpart, it is almost as effective. Considering low-power requirement for the pro-pose system appears more attractive and practical in real structures.