Self-Adaptive Control

Introduction

The Self-Adaptive Control is one of the major types of adaptive control. It is comparable to feedback compensation because the adaptation of controller parameters is based on the measurement of closed-loop performance and aim to optimize it by using a reference model of the process or by estimation techniques.
In Self-Adaptive Control, the controller parameters are derived for every process conditions and it is not programmed as in the case of Programmed Adaptive Control. The parameter adjustment loop search for the optimum values for the controller parameters on-line.

Types of Self-Adaptive Control
The two mechanisms that come under self-adaptive control are,
1. Model Reference Adaptive Control (MRAC) and 
2. Self Tuning Regulator (STR).

Model Reference Adaptive Control (MRAC)

Model Reference Adaptive Control (MRAC)

When the process parameters are unknown or change in time, in order to achieve and to maintain the desired performance, an adaptive control scheme known as Model Reference Adaptive Control (MRAC) is implemented. Here, a reference model which is the realization of the process with desired performances. This scheme is based on the observation that the difference between the output of the process and the output of the reference model (called as the process-model error) is a measure of the difference between the real and the desired performance. The Adjustment Mechanism directly adjusts the parameters of the controller in real time in order to force the process-model error to zero. MRAC is also known as Direct Adaptive Control because of the direct adjustment of the controller parameters using the model output.

Self Tuning Regulator (STR)

Self Tuning Regulator (STR)

The basic idea of STR is that a suitable controller can be designed on-line if the process parameters are estimated on-line from the available input-output measurements. The process output and the estimated or predicted output are compared. The adaptation mechanism uses the error between the process output and predicted output (called as prediction error or process-model error) to adjust the controller parameters in such a way to minimize the prediction error under the specified conditions. STR is also known as Indirect Adaptive Control because the adaptation of controller parameters is done by the on-line estimation of process parameters.

Also read other control schemes

Inferential Control 

Programmed or Gain Scheduled Adaptive Control

Adaptive Control - An Introduction

Override Control

Feedforward Control

Feedback Control

Cascade Control

Ratio Control

Feedforward control vs Feedback control