Theses supervised by Doç. Dr. Tolgay Kara
10 theses · Gaziantep University
Exponential reaching law based adaptive sliding mode control of multivariable liquid level processes: An experiment study on quadruple tank process control
Industrial processes are complex and have uncertainties. The dynamic behaviour of these processes are needed in order to analyse and effectively control the process outputs of the system. An example of such a system is a Quadruple Tank Process (QTP) with multiple inputs and multiple outputs and nonlinearity due to coupling of the tanks. Therefore, Exponential Reaching Law Based Adaptive Sliding Mode Control (ERL-ASMC) has been proposed to solve the problem of nonlinearity, disturbances and parameter uncertainties. It is able to adjust to the control gain of the discontinous switching function so as to reduce the reaching time. It is very useful in minimizing chattering due to high frequency gain. It has also been compared with Adaptive Sliding Mode Control (ASMC) and Adaptive Pole Placement Control (APPC) which also use Parameter Estimation method like the ERL to determine the parameters of the QTP. These controllers are then finally compared with a Proportional-Integral-Derivative (PID) in order to better evaluate their performances. The evaluations are carried out based on the transient response specifications and the performances indices. The controllers are first simulated and then implemented on the real QTP in the laboratory. The proposed ERL-ASMC has the best performance in simulation and satisfactory results in the experiment. Key words: Quadruple Tank, Parameter Estimation, Adaptive Pole Placement control Sliding Mode Control, Exponential Reaching Law Based Adaptive Sliding Mode control
Modeling, simulation and control of fuzz recycling process for textile industry
The importance of recycling comes from the benefits it returns to the environment. Pointing on carpet manufacturing, after the production process is completed the carpet passes through "thread trimming" stage, which results in fine threads (fuzz). The fuzz is disposed only because it is difficult to reuse it in current physical form. The current study focuses on the mathematical modeling and simulation of a fuzz recycling machine that helps reform the fuzz into a reusable solid material. The process is divided into two stages, filling the vessel with fuzz and compressing it until it reaches the desired density. Filling process modeling relies on indicating the change in fuzz concentration through the pumping procedure, which leads to a relation expressing the effect of the fuzz concentration variance on both released air and vessel density through this stage. Compression process modeling counts on a compressor that applies specific pressure on the air-fuzz mixture to release air through filtered exit holes and compress the fuzz to produce it in solid form. Using calculations and available data it is found that the relation of volume and density of the material during this process is linear and simulating the model has shown the effect of pressure on both volume and density. Additionally the pressure dynamics modeling the pressure change in each stage is shown. This system is expected to help create a new recycling machine that can save large amounts of carpet industry wastes.
A simulation study on impedance control of two-degree-of-freedom robotic manipulators
In this research, motion control of a two-degree-of-freedom robotic manipulator with rigid links is studied. Robot control with this approach needs a control system that includes dynamic models of applied torques effecting the positions, speeds, and forces at joints. Hybrid impedance control is proposed in this research to address this need. For hybrid impedance control to be fully understood, position control, force control and impedance control methods are discussed in depth. In the light of the literature review, the mathematical kinematic and dynamic equations of two-degree-of-freedom robotic manipulators are first derived. The manipulator model is converted into a simulation model in computer simulation environment, and the accuracy of the model is proven through open loop tests. The force and position of the end effector of the robotic manipulator is computed with a hybrid impedance control model that can be controlled simultaneously. The designed hybrid impedance controller is integrated into the created robotic manipulator model, and integrated into the closed-loop simulation model along with the PID control system with the purpose of performance assessment. The simulation results are presented in graphs and compared with the desired force control, position control, speed control and joint angle with the force control, actual position control, velocity and joint angle. The simulation is continued until the margin of error reaches the minimum level. Hybrid impedance control is determined to have superior performance compared to other control methods tested. Key Words: 2 DOF Robotic Manipulator, Hybrid Impedance Control
Design and simulation of a PID neural network controller for PMDC motor speed control
Direct current (DC) motors have many difficulties when controlling angular velocity in a variety of applications. The perfect controller cannot be carried out by traditional control alone due to the nonlinear properties of DC motors, design constraints, and mechanical variation caused by the operation conditions. This study proposes a design for an artificial neural network-based PID controller (ANN-PID) to control the speed of a permanent magnet DC motor (PMDC) in two methods. A detailed analysis is performed based on the simulation results of both methods. The proposed controllers are numerically simulated for various test conditions including; set-point changes, step changes in the load torque, and parameter variations, then the suggested techniques were compared in a comparative study with a traditional PID controller based on the transient response specifications and the performance indices to validate the performance of the controllers. The simulation results demonstrated that the controllers have improved dynamics, static performance, and less overshoot. The methods described here achieve control more effectively than the conventional control approaches under both nominal and disturbed test conditions over different operating ranges.
Elektrikli tekerlekli sandalyenin geri beslemeli hareket kontrolü
This thesis presents the design and implementation of feedback control for the direction and speed of an electric wheelchair through the appropriate feedback signal. Wheelchairs and electric wheelchairs present various challenges for individuals who use them. In this research, a new control system design for an electric wheelchair is presented to feedbacks on the data obtained. Instead of the standard motor driver of the electric wheelchair, it was thought to make a new and economical motor driver. An encoder and motion controller are added to the control system architecture for precise motion control. All feedback signals are evaluated and codes are written for different possible scenarios. The design scenarios of the system where different algorithms are applied for smooth driving in case the electric wheelchair traveling on the determined route returns to the desired position when it is not hit from the left or if the driver is overweight or underweight are examples of design scenarios. As a result, our findings show that the proposed design can control the autonomous system in a closed loop. In addition, tests on a real wheelchair prove the effectiveness of the model and the designed control system. The results of the study form the basis for the ongoing autonomous wheelchair project and contribute to the research field of assistive technologies for the disabled and the elderly.
Development and testing of a graphical simulator for wastewater treatment plant design and analysis
In this study, a computer based graphical interface is developed to improve wastewater treatment systems. Depending on the desired input-output relationship of the system, the simulator is run to show the possible results for each step of the processes of a wastewater treatment system. In order to describe the treatment performance on the wastewater, each subsystem responses, which are predefined and mathematically modeled, are observable. It is possible to observe the system behavior on maximum and minimum values of physical, biological and chemical wastewater components. The subsystems are constructed elastically; the simulator allows adding different mathematical responses to the subsystems on different chemical and physical components. The organic-inorganic and chemical-physical responses of the system are simulated distinctly. The wastewater treatment characteristics for different components of any type are planned, observed and optimized. All relevant variables such as depth of pools, flow speed and sewage characteristics can be changed on the simulator. Whale Optimization Algorithm (WOA) is used to achieve the desired optimized solutions. The simulator is run for the Gaziantep Oğuzeli WWTP conditions. The results of whole simulator are concluded with both economic and environmental trajectories depending on the desired outputs and the control mechanisms of the system. Moreover a Unified Tank Model (UTM) for the aeration and sedimentation phases is proposed and investigated as part of this study. Environmental aspects are taken in to account for all steps of this study. Keywords: Wastewater, Wastewater Treatment, Mathematical Modeling, Simulation, Control.
Robust adaptive type-2 neural fuzzy sliding mode control of a class of nonlinear systems
This study aims to develop an adaptive control scheme for the control of a class of flexible multi-body nonlinear systems with infinite dimensions and extremely coupled dynamics. The finite-dimensional model is obtained by the assumed modes method (AMM), but there are uncertainties in the truncated model, which makes the system a challenging control problem. The proposed adaptive control scheme is the hybrid of the sliding mode control (SMC) and the type-2 neural fuzzy system (NFS). A newly modified conjugate gradient (CG) algorithm is used to optimize the NFS parameters, thus enhancing its self-adapting capabilities. The control law of the proposed control scheme requires the estimation of the unknown system functions, which is provided by the adaptive NFS. The stability of the control scheme is guaranteed by the Lyapunov stability theorem. Several other intelligent control schemes have also been tested to provide a comparison with the proposed control scheme. The simulation results clearly show that the proposed control scheme improves tracking efficiency while managing the inherent deflections of the system, making it an appropriate adaptive control technique for the class of flexible multi-body nonlinear systems.
Supervisory monitoring and control of temperature and humidity using a wireless sensor network
The wireless sensor networks (WSNs) are contemporary techniques that can provide several options to users by controlling and monitoring devices according to reception data from numerous sensors. The main goal of this research is to implement and test supervisory control of temperature and humidity values in a WSN configuration. This thesis presents a simulation model and test results of the WSN and control system in computer environment as well as the experimental results and control implementation in real time. The experimental system consists of a DHT11 sensor, ATmega-328 microcontroller and software for control of the variables. In this research ZigBee protocol is preferred for its suitability for long-distance and speed to transmit information with good performance to implement WSN. A Proportional Integral Derivative (PID) control law is designed and tuned for temperature and humidity control loops in the WSN. Performance of the designed WSN and control system is tested via computer simulations and real time tests, and results reveal satisfactory performance proving applicability of the proposed system in real applications.
Sliding mode anti-sway control and disturbance rejection of gantry crane
In this study, a robust sliding mode control (SMC) strategy is suggested for suppression of load sway and rejection of disturbances in gantry cranes. First the dynamic model of gantry crane is derived from the general laws of motion and Lagrangian energy analysis. Then, a proper disturbance rejecting robust SMC mechanism with anti-sway property in conjunction with a Linear Quadratic Regulator (LQR) design is developed. Suggested control system is tested via computer simulations along with traditional Linear Quadratic Regulator (LQR) state feedback control and LQR with Proportional Integral Derivative (PID) control methods. The control simulation results of the crane dynamic model under ideal conditions without friction and actuators is compared. Simulation results reveal that suggested SMC strategy exhibits superior performance in suppressing sway and rejecting disturbances in gantry crane control.
Identification of mechanical backlash in electromechanical systems for control purposes
In the last half century, especially industrial, robotic and automotive sectors started to take an important place in our lives with developing technology. The working principles of these sectors are mostly based on mechanical systems. One of the biggest factors that affects the working mechanism of a system negatively is the backlash formed in the mechanical parts. In order to reduce the damage caused by the backlash in mechanical systems, backlash must be compensated. The first stage of this starts with the identification of backlash size. Therefore, this thesis focuses on the identification of backlash size formed in mechanical systems. For this purpose, using computer simulation two feedback control system models, the first one for a mechanical system with backlash and the other one for a mechanical system without backlash, have been created. Kalman filter has been designed to see and understand the effect of backlash on load positions more accurately. Changes in load positions have been observed by using step input signal. Sum of the absolute value of the differences between load positions with backlash and load positions without backlash has been obtained. In order to achieve the value of estimated backlash size, an equation proportional to the obtained absolute sum result has been formulated. Estimated backlash values have been obtained by performing simulation experiments for different actual backlash values and their accuracy has been confirmed.