Implementation of Embedded Soft Sensor for Bioreactor on Zynq Processing System

Abstract

At Present, developing a complex algorithm using an embedded system, particularly in a hybrid computing system, has obtained vital importance in many fields. This paper presents the design and developing a soft sensor for a bioreactor using Zynq processing system (PS). In this work the Extended Kalman filtering (EKF) technology is used for the design of a nonlinear model which is used to estimate the state of lactose concentration of a simulated bioreactor. In this work a base design for the Zynq Processing System is created in VIVADO and the Zybo board is used as hardware platform. Then the non-linear designs are developed in C as software application which is made to run on Zync Processing System. A comparative study is made on computation time and consumption of resource between the PC based realization and Zynq Processing system realization. The hardware implementation of the soft sensor for measuring lactose concentration of a Bioreactor using Zynq PS is faster than PC based realization with a speed up of 1.1 times when compared to that of PC based realization. The above parameters are achieved by using an average of 8% of all resources.