Analysis on Data-Based Integrated Learning Control for Batch Processes - Intelligent Computing for Sustainable Energy and Environment - page 131

uncertainty and disturbance exist. Therefore, the integrated optimization control tech-
nology is required in order to derive the maximum benefit from batch process, in
which the performance of time-axis and batch-axis are both analyzed synchronously,
such as the works done by Amann, Gao, Lee, Xiong, Rogers, Kurek and et.al [3-17].
Motivated by the previous works, an integrated learning control system based on
input-output data is proposed in our previous work. Based on that paper, the conver-
gence and tracking performance of the proposed integrated learning control system
are firstly given rigorous description and proof in this paper.
The paper is structured as follows. Section 2 presents the proposed data-driven
based integrated learning control system. Section 3 presents performance analysis.
Simulation example is given in Section 4, followed by the concluding remarks given
in Section 5.

2 Data- Based Integrated Learning Control System Design for

Batch Processes

The proposed integrated learning control system consists of: the iterative learning
control (ILC) working as feedforward controller and adaptive single neuron predictive
controller (SNPC) playing as feedback controller. For the convenience of discussion,
the number of batch and batch length are respectively defined as k and f

which is

t

( ) ( )

divided into T equal intervals.

ILC ukt

, and u k t , are ILC control variable and

SNPC

( ) ( ) ( )

SNPC control variable of time t in k -th batch,

y

ukt , = u kt , + u kt ,

ILC SNPC

,

( )

is the targeted end-product quality,

ykt is the corresponding product quality of

,

( )

two control actions,

ykt is the predicted output of data-based model. Since batch
process is repetitive in nature, the model prediction at the end of the k-th batch,
( )

ˆ ,

( )

ykt can be corrected by ˆ , f  ykt , f . During k -th batch, the control policy of U

, ILC k

obtained from ILC optimization controller and the control policy of

U computed
from SNPC controller are summed as U and is sent into batch process to improve
the performance, Y and ˆ Y are respectively product quality variables and predicted
product quality variables. As discussed above, the proposed integrated learning opti-
mization control action can be described as

, SNPC k

UU = + U (1)

k ILC k , SNPC k ,

( ) ( ) ( )

(2)

ukt , = u kt , + u kt ,

ILC SNPC

( )
( )

low up

u ≤ u k t ,

,

≤ u

low up

y ≤ y k t ≤ y