Deployment of Cloud-Based Wireless PI Controller for Non-Linear Delayed System

Liquid level regulation is important in a wide range of process industries like petroleum refineries, chemical production plants and water treatment plants. Spherical tanks show a high nonlinear behavior because of the continuous change in cross-sectional area. Hence, their modelling and control require more effort. Efficient management of liquid levels in nonlinear systems is important to guarantee the safety of operations, quality of produced products, and efficiency of any process. This paper presents the design and implementation in real time of a wireless classical Proportional- Integral (PI) control of a non-linear Conjoined Tank Liquid Level System (CTLLS) via LabVIEW. A wireless sensor network architecture is being used, where the process variables are being collected with the help of distributed sensor nodes and being sent to a realtime gateway for the purpose of monitoring and control. The settings of the PI controller are determined through classical tuning methods, the Ziegler-Nichols (ZN) method and the Skogestad (SG) method based on experimentally determined process models. A robust data logging and communications framework is installed. Liquid level, set point, error, and time stamp information are recorded locally continuously as well as sent via a Microcontroller with Wi-Fi setup to a cloud platform. This allows remote access, visualization, and post-processing of system performance studies, thus improving dependability and scalability. Experimental results show that the proposed wireless PI control strategy is effective in controlling the liquid level in the nonlinear spherical tank system in both the nominal and disturbed operating cases.