In the process industry, Coupled Tank system is used for liquid level and flow rating control. Coupled tank is a plant model that related to process industry for example chemical process. In this plant model level controls are very important because, to get an accurate result is hard. Coupled tank has two tanks which is flow to the first tank and then will flow to another tank. Because of this problem, the advanced controller Linear Quadratic Regulator (LQR) has to design to control the flow and get an accurate result.
By using computer as the main control, for tank one or both tank user can control the level of liquid in the basic concept that the coupled tanks system works. Advanced controller is needed to be implemented to control the liquid level automatically. There have a device that connected to communicate between computer and the plant. This project focuses on the modeling for coupled tanks system as a plant and designs the LQR controller with implementation of real-time. By using the guide manual book for the plant system which is has Simulink with MATHLAB for the experiment. Visual Basic is used to get the value of state feedback that produce by LQR to see the response in real time process based on the simulation result.
Introduction
In the process industries, the control application of the liquid level and flow rating between tanks is a basic problem and is widely use on control system especially in Chemical industries. Coupled Tank liquid level system is one example of the application of industry liquid level control systems plants that normally are used in industry.
This system is usually being controlled by advanced controller such as PID, Fuzzy logic control and other advanced controllers to control the system. Liquid level are implementation of Linear Quadratic Regulator (LQR), this is one of the new methods of controlling that will be discover in this project on Coupled Tank liquid level system. Theoretical of optimal controls that operating in dynamic system at minimum cost is called LQR. Software CTS-001 has been use in computer will control the plant which is a coupled tank system with implementation of LQR controller through this project. CTS-001 is a model that represent as a coupled tank and basically, software will developed using data acquisition card (DAC), LABWINDOWS or CVI environment.
By using this method of CTS-001 transient response, steady state error analysis, and also controlling tuning method are real-time implementation. To optimize the efficiency of the system such as accuracy, the implementations of LQR controller are the main aim of this project on Coupled Tank water level system.
Problem statement
Control Algorithms are involved real-time implementation control to control a certain process like liquid level and flow rating. In terms of implementation in real-time performance and each control features, control level of a coupled tank is chosen. Linear Quadratic Regulator has been applying as a controller for this project, real-time implementation will be used to controlling liquid level process. To control of fluids level in storage tanks and chemical blending, it is a common problem control in process industries. Fluid will be supply at a constant desired liquid level as a constant rate with regulated flow of liquid into and out have to achieve. By using various techniques to compensate, control requirement has to be implemented with the control algorithms. There has they own advantage and disadvantage.
The most commonly controller that has been used to control liquid level system is Fuzzy Logic Controller and PID controller in coupled tank system. Wide arrays of other control techniques have been applied to meet the control objective of the system. For set point tracking and load disturbance, effects of conditions and uncertainty reducing, time response behavior such as stability, rise-time (Tr), overshoot (OS%) , and steady state error (ess) are various factors which considered for designing in the controllers.
Objective (s) of the Project
This is the objective of this project is: -
To model the plant of coupled tank liquid level system with the implementation on real time.
To design an advanced controller for the plant, coupled tank liquid level system using LQR controller.
Study and understand the implementation of LQR controller to see the performance of liquid level control on coupled tank liquid level system.
To get the simulation of LQR controller for liquid level control by using MATLAB Simulink (Second-Order system).
To analyze the performance, accuracy and stability of liquid level control comparison between experiment and simulation result.
Project Scope
Coupled Tank liquid level system will be Implement LQR controller using Visual Basic 6. By considers corresponding to water level or height, control water flow into the tank to another tank is a method that has been applied. Second order system (SISO) is used for this application. Behavior and outstanding characteristics research on LQR had been discussed to see prior implementation in industry.
Hardware
a. Design coupled tank plant.
b. Interfacing DAQ card with computer.
Software
a. Simulation the system using MATLAB
b. learn and study Design GUI using Visual Basic 6
Literature Review
This part is review research on couple tank system which considers controller linear quadratic regulator and direct digital control. To controlling the liquid level in the coupled tank system is a method that has been control. Optimal control techniques concern as linear quadratic regulator (LQR) is one of the dynamical system and control input to make the optimal control decisions.
Coupled Tank
In the process industry, the control of the liquid level and flow rate between tanks is a basic problem. The level of the fluid in the tanks must be controlled. Such as, liquid to be pumped then store in the tanks after that pump into another tank. This is the technic that the process industry required [1]. This process is commonly happen for chemical industry which the liquids is processed by chemical or mix some treatment in the tanks, that is why the fluid has to be controlled and also the flow between tanks must be regulated. Situation of both tanks are coupled together, so that the level of fluid n flow interacts and this has to be controlled. The performance and the significance of the controlled system are investigated under variation in system parameters and also in presence of an external disturbance [2]. In chemical engineering systems they need to control the level and flow in tanks.
Linear Quadratic regulator
Complex systems that have stringent performance requirements are for optimal Linear Quadratic Regulator. LQR is also a powerful method to find a controller over the use of the other controller, and it is more robust to the parameter variations also easy to meet the requirement [3]. MATLAB can solve LQR problem must be solved via a Computer-Aided-Design (CAD). By using the CAD packages optimization problems has solving challenge lies in the weighting matrices are chosen [4]. Consider systems effected by random bounded nonlinear uncertainty so that classical optimization methods based on linear matrix inequalities cannot be used without conservatism also is a blend of randomization techniques for the uncertainty together with convex optimization for the controller parameters.[5]
System Identification
Basically the second order single input and single output (SISO) system is used for Coupled tank System. The figure (1) of the block diagram below show for an open loop system in second order (SISO) with interfacing, and block diagram (2) of an open loop system is in second order (SISO) with LQR controller.
Methodology
This is the plan for provide further details of methodology and approach by completing this project. A control system variable is one of the level controls which this is very important in the industries. Interacting tanks in the actual industries as low-cost pilot plant by using AISB Coupled-Tank Control Apparatus CT-001. Real-time implementation has been applied to the CTS-001 to be used.
Visual Basic is involved modeling in the system, such as system identification of non-parametric model. The system of the second order single input single output (SISO) is model to obtain the specialized for both tanks system. Discussion for the process plant, software and data acquisition card will be implementing throughout by flow chart of this project. Based on this flow chart, it shows the progress of this project.
Basically, analysis of the problem will start first and follow by defining hardware and software. It has been divide into two parts that is software and hardware. The hardware is the coupled tank which is a plant of this project setup. The installation or connection between Advantech 4716 USB DAQ and the plant is communication for testing and calibration. The software part is the process which includes modeling n simulation the system in MATLAB, installing the Visual Basic 6, designing GUI on Visual Basic 6. For conclusion for this project, the troubleshooting process for hardware and software is done. Refer the flow chart above of this project for further process.
Expected Results/Benefit
Linear Quadratic Regulator controller (LQR) is more robust to the parameter variations and easy to meet the requirement. The result shows that LQR produced better response compared to conventional PID control strategies in time domain. LQR has a good performance in many application because it reliable and accurate.
Expansion of knowledge - Gain knowledge about Linear Quadratic Regulator (LQR) advanced or modern
Controller
Project Publications?
Specific or Potential Applications: - Industries of Petro-chemical
Industries of Water treatment
Milestones
Milestone 1 - Get the title for project. Do research and find related case study such as journal.
Milestone 2 - Determination of hardware and software between plant and controller.
Milestone 3 - Experimental on hardware (plant), coupled tank in the laboratory. Do the basic
Experiment based on manual given.
Milestone 4 - Testing and do simulation. Get the data.
Milestone 5 - Find and do the research on advanced controller that can improve the performance better
than conventional controller.
Milestone 6 - Assessment and analyze the data from plant and conventional controller.
Milestone 7 - Writing report and proposal.
