Browsing by Author "Dereje, Shiferaw (PhD)"
Now showing 1 - 20 of 36
Results Per Page
Sort Options
Item Comparative Analysis of PID and Fuzzy Logic Controller for Induction Motor Speed Control(Addis Ababa University, 2019-10) Awole, Hussen; Dereje, Shiferaw (PhD)Induction motor (IM) is the most rigid, and relatively less expensive machine but much difficult to control. The advent of field-oriented control (FOC) makes IM useful in variable speed drive applications. The concept of FOC is to separate the torque and flux producing current and then control the torque and flux separately. The advent of different control theory makes difficulty in the choice of an appropriate controller. In this thesis, a comparative analysis of fuzzy and PID control for IM speed control has been done. To solve this problem first an indirect field-oriented control (IFOC) method motor control is designed. In this design, the direct current ������ is kept constant for a fast response. In addition, the motor is modeled using rotor flux and stator current as a state variable. This model is very important due to the presence of measurable quantity (stator current), and to mathematically quantify the alignment of rotor flux on the d-axis. Both PID and fuzzy control of IM has been verified using simulation on MATLAB/SIMULINK. The performance of both PID and FLC is analyzed in terms of reference tracking, load variation, parameter variation, low-speed tracking, and speed reversal. The PID controller results 0.3s settling time with 10% overshoot and the fuzzy controller 0.2s settling time with 0% overshoot.Item Design and Analysis of Fuzzy Logic Based Controller for Flow and Level Control of Cane in Wangi Sugar Factory(Addis Ababa University, 2018-06) Yohannes, Solomon; Dereje, Shiferaw (PhD)In a sugar production, flow and the amount of cane fiber carried by cane carrier varies due to non-uniformity of cane supply. The continuous variation of cane fibers flow and the level of cane fiber in chute during the cane juice extraction inversely affect the cane juice extraction efficiency of mill. In this thesis we have developed algorithm for a three input fuzzy controller with an aim to maintain the cane level in chute and flow during cane juice extraction. The developed controller generates signal that required controlling cane carrier motor speed depending upon the value of cane level in chute, quantity of cane on rake carrier and flow rate. The three inputs fuzzy controller is developed and simulated for six cases by using fuzzy logic toolbox of MATLAB. The performance of the controller is compared in terms of disturbance rejection, transient and steady sate performance. It is observed from the simulation results that the average overshoot is 0%, rising time is 0.0817 seconds and the settling time is 0.274 seconds with the proposed fuzzy controller while overshoot is 7.62%, rise time is 0.0513 second and settling time is 0.16 seconds with PID controller. Moreover, the robustness and disturbance rejection of the controllers is checked by parameter variation like time constant, delay time & DC gain and giving disturbance signal after settling time respectively. It is further observed that the proposed controller has better disturbance rejection and more robust.Item Design and Comparative Analysis of Genetic Algorithm Tuned Fractional and Integer Order PI Controllers with Adaptive Neuro fuzzy Controller for Speed Control of Indirect Vector Controlled Induction Motor(Addis Ababa University, 2019-01) Girma, Kassa; Dereje, Shiferaw (PhD)This thesis presents design and comparative analysis of fractional order PI controller, integer order PI controller and an adaptive neurofuzzy controller trained by input output data from fractional order PI controller for indirect vector controlled induction motor. The parameters of the two PI controllers were genetically optimized using square of error as a fitness function. The proposed neurofuzzy controller trained with input and output data of fractional order PI controller incorporates fuzzy logic algorithm with a multilayer artificial neural network structure using hybrid learning algorithm. This improves the performance of induction motor drive. The fractional order model of induction motor has been also investigated using simulation results and it was inferred that optimized model of induction motor is an integer order model. The performance of adaptive neurofuzzy inference system controller, was compared with fractional and integer order PI controllers using MATLAB simulation results with different operating conditions. It was observed from the simulation results that by using ANFIS, FOPI, and IOPI controllers, for the reference speed of 50 rad/sec, the percentage peak overshoots were 0.496%,13.068% and 15.698% respectively. Thus, ANFIS shows dramatic decrease in overshoot. Also the speed reaches its desired set value at 0.15 second in ANFIS controlled IM drive. These show the effectiveness of the designed neurofuzzy controller and the designed neurofuzzy controller tries to speed up the performance of IM drive. On the other hand, FOPI controller showed better performance than IOPI controller for IM drive, this is because of FOPI controller has one additional parameter for tuning which is integration order.Item Design and Compare Adaptive Neuro-Fuzzy Inference System (Anfis) With Genetic Algorithm (Ga) Tunned Pi Controller for Speed Control Of Vector Controlled Induction Motor Drive(Addis Ababa University, 2018-11) Daniel, Arega; Dereje, Shiferaw (PhD)Nowadays, vector controlled induction motor drives with variable speed applications are widely used in order to achieve good dynamic performance and wide speed control. The conventional speed controllers for vector control of induction motor drive suffer from the problem of stability; besides, these controllers such as PI/PID controllers show either steady state error or sluggish response to the agitation in reference setting or during load perturbation. In this thesis a new method of controlling technique based on the combination of Artificial Neural Network (ANN) and fuzzy logic (FL) is proposed to improve the speed control of indirect vector controlled induction motor drive. Indirect vector controlled induction motor with genetic algorithm (GA) optimized PI controller is developed and is replaced with adaptive neuro-fuzzy controller (ANFIS) to overcome the problem of overshoot occurred in PI controller and to obtain quick steady state response and better speed control. The proposed technique is implemented using MATLAB/Simulink. In this thesis, the speed, torque and stator current responses with GA based PI controller and proposed adaptive neuro-fuzzy controller are compared and found that the proposed ANFIS based controller showed increased dynamic performance. The proposed adaptive neuro-fuzzy controller is better in overshot which is 0.475% and that of PI controller is 14.368%, raise time and settling time.Item Design and Implementation of Feedback Linearization based Adaptive Stabilizing Controller Coupled with Fuzzy Logic Swing-up for Pendulum on a Cart(Addis Ababa University, 2019-11-19) Daniel, Abebe; Dereje, Shiferaw (PhD)This thesis address an adaptive stabilizing controller for inverted pendulum on a cart based on feedback linearization coupled with an adaptive fuzzy logic based swing up controller. First feedback linearizing control signal is derived by decomposing the system into cart subsystem and pendulum subsystem. Then adaptive inverse control technique is applied to each feedback linearizing control signals. An adaptive inverse control method is used for compensation of unknown parameters of an inverted pendulum on a cart, while feedback linearization is used to cancel non-linearity in the system. The pendulum is driven from it's pendant position to inverted position using an adaptive fuzzy logic based swing up controller. When the pendulum reaches near it's inverted position, the stabilizing controller takes over the swing up controller. The MATLAB/SIMULINK simulation shows that the proposed controllers adapt to unknown mass of a cart between 0:1kg - 4kg and mass of a pendulum between 0:01kg - 4kg. The performance of the stabilizing controller on hardware experimentation under unknown mass of a cart and mass of a pendulum shows that the proposed controller is a solution to inverted pendulum stabilization problem.Item Design and Simulation of Mobile Pick and Place Robot Manipulator (MPPRM)(Addis Ababa University, 2020-02) Kalkidan, Yirmed; Dereje, Shiferaw (PhD)Self-balanced two-wheeled mobile robot manipulators (TWMRMs), which is based on an inverted pendulum system, is dynamically (for handling a payload) and statically unstable. While different configurations of self-balanced two-wheeled mobile robot manipulators (TWMRMs) exist, the workspace of these systems is restricted by their current configurations and designs. This study presents the performance of using a fuzzy-PID control algorithm as part of a PID control scheme for the control of 5 DOFs self-balanced TWMPPRM that could be useful for industrial applications such as pick and place and materials handling hazardous products in the chemical manufacturing industry. The system under investigation offers solutions for industrial robotic applications requiring a limited working space. The non-linear mathematical model system derived from Lagrangian modeling approach is simulated in the MATLAB / SIMULINK framework. Fuzzy-PID control of a decoupled nature is designed and implemented. Various operating scenarios with several initial conditions are used to test the robustness and performance of the system. Using integral absolute error (IAE) strategy for payload free movement, it is observable that the system takes around 2.2870 seconds to settle by implementing the PID control strategy, which is greater than the settling time (0.7800 seconds) of the Fuzzy-PID control, and for peak time and rise time it takes 0.5710s and 0.2790s respectively for PID. Furthermore, for peak time and rise time it takes 0.4400s and 0.2300s respectively using Fuzzy-PID control strategy. The simulation results showed the effectiveness of the Fuzzy-PID control approach in improving system performance compared to the PID control system.Item Design and Simulation of Speed Control of PMSM with Fuzzy Logic SelfTuning PID Controller using MATLAB(Addis Ababa University, 2017-10) Henok, Teshager; Dereje, Shiferaw (PhD)This thesis focuses on the speed control of Permanent Magnet Synchronous Motor (PMSM) drive system with Fuzzy logic self-tuning PID controller using MATLAB/ Simulink 2015a. The performance of the proposed controller of PMSM was analyzed in terms of speed tracking capability, torque response quickness, high and low speed behavior, parameter sensitivity and speed reversal conditions. Comparison is also made between FPID with the conventional PI and PID controller mechanisms. The result shows that the FPID has more significant advantage in overshoot reduction by (11.7% and 21%) than PID and PI controllers and with very good steady state error of 0.035% without load test. While with load test the FPID controller have good performance with 0.043% steady state error percentage.Item Design of Fractional Order PID & Neural Network Controller for Magnetic Levitation Train System(Addis Ababa University, 2018-12) Yibeletal, Kassahun; Dereje, Shiferaw (PhD)Magnetic levitation systems are systems which operate based on the principle of magnetic attraction and repulsion to levitate an object and widely used in frictionless bearings, high-speed Maglev passenger trains, levitation of wind tunnel models, and so on. Their attracting feature is that they have a zero frictional force which considerably reduces the energy required to drive the systems. However, magnetic levitation systems are highly nonlinear and open loop unstable which makes their control very difficult. Moreover, magnetic levitation trains are more advantageous than conventional trains because train weight is lower due to the absence of wheels, axles, and engine, and hence can save huge energy consumption; the energy loss due to unwanted friction is negligible; it allows the train to move at a very high speed and to be environment friendly. This Thesis work investigates the control mechanism of Maglev systems based on linear unity feedback controller (PID and FOPID) with Taylor series linearization techniques and nonlinear controllers (ANN) based on nonlinear methods. To improve the closed loop performance of the PID controller more advanced PIλDδ controller is used for the system. Stability is also ensured due to the additional tunable parameters λ (0.99811) and δ (0.99998) respectively. Then the proposed FOPID controller has resulted good performance i.e. the PI��D�� controller improved a design performance specifications of settling time, overshoot, steady state error and peak response from (0.103 to 0.015 second), (18.452% to 4.37%), and (0.18 to 0.043) and (1.184 to 1.043) cm in Y direction. Lastly, for position control and stabilization of the Maglev train system, a powerful ANN controller is designed. Furthermore, a comparison between simulation results of the Conventional PID, fractional order PID, and NN controller is made to check the performance characteristics and stability of the Maglev train system. FOPID controller is more improved by ANN controller; performance characteristics of Maglev train obtained by ANN controller has resulted good. The neural network controller improved rise time, settling time, overshoot, steady state error and peak response from (0.009 to 0.0057 second), (0.015 to 0.0069 second), (4.37% to 0.864%), (0.043 to 0), and (1.043 to 1) cm in Y direction.Item Design of Fuzzy Sliding Mode Controller for Heartbeat Pacemaker Based on ECG Signal Tracking(Addis Ababa University, 2020-11) Anteneh, Nebyu; Dereje, Shiferaw (PhD)Every year, heart disease is becoming the major cause of death. Therefore, the need arises to find advanced approaches to keep the patients safety. Implantable cardiac pacemakers are an electronic device that can track and boost the heart rate and manage rhythm disorders. This thesis presents nonlinear control of heartbeat model. Because of certain severe cardiac arrhythmias display nonlinear feature which is usually correlated with unpredictable and oscillatory behavior, a nonlinear technique is used to model heart electrical activity. Zeeman’s heartbeat model was used to generate ECG signals. Existing model of heartbeat was analyzed and revised by integrating the control input to add the control mechanisms as a pacemaker. In this study, sliding mode control (SMC) is applied to heartbeat model in order to track and generate real ECG signal. Fuzzy logic algorithm was also used with SMC to reduce chattering happened due to high frequency oscillations around the sliding surface that will shorten the life span of pacemaker. Therefore, a fuzzy sliding mode controller (FSMC) for cardiac pacemaker based on ECG signal reference tracking system was designed. In addition, since the heartbeat pacemaker is disturbed by the brain signal and sensor output delay, the robustness of the system to disturbance, parameter variations and possible time delay on the feedback system were analyzed. The effectiveness of the proposed method was verified through simulation studies using Matlab/ Simulink software. The proposed control law has shown satisfactory performance in terms of tracking ECG signal of the actual data, obtained from the MIT- Boston’s Beth Israel Hospital (BIH), and the physioNet database by eliminating chattering compared with the use of SMC controller. In addition, the result of root mean square error was reduced by 0.00005 % and total harmonic distortion analyzed from FFT window was reduced by 121.35 % when FSMC was applied to the system compared with SMC. The control strategy was also found to be robust with respect to external disturbances, parameter variations and random feedback delay. Therefore, control algorithm will be applied in dual sensor cardiac pacemakers for clinical use.Item Design Of Intelligent And Hybrid Based MPPT Controller For Photovoltaic Water Pumping System(Addis Ababa University, 2018-12) Abemelek, Molla; Dereje, Shiferaw (PhD)Maximum power point tracking plays an important role for Photovoltaic power pumping systems because it optimize the power output from a Pholovoltaic system for a given set of conditions. This thesis presents a maximum power point tracker using artificial intelligence and hybrid methods for a standalone Photovoltaic water pumping system. This work focused on designing of different intelligence control methods to get maximum amount of power for Dolo Ado Woreda refugee camp appropriate power for the required amount of demand to pump water from the ground to 600 residential house hold needs. In this work scaling and sizing the whole components of the standalone photovoltaic water pumping system, such as, Photovoltaic panel, direct converter, inverter, and low pass filter is applied to generate a 29.04kW power by using a boost converter as a supporter for maximum power point traking algorithm by adjusting the duty cycle of the boost converter to maximize the output to the inverter which is feeding an alternative 17.5kW current load would achieved with three proposed intelligent controllers. After sized and designed the proposed system components, each System elements are individually modelled in MATLAB/SIMULINK and then connected to assess performance under different environmental conditions. First, each technique is compared with the direct connection matched system. The results show that the direct connection PV system response oscillates far from the tracking point and the three proposed controller method dynamics responses are around the maximum power point under different level of temperature and irradiation. The performance of the three proposed controllers for photovoltaic water pumping system is evaluated through simulation studies and compared. The simulation results show that the efficiency of the photovoltaic water pumping system with the fuzzy logic, artificial neural network and artificial neuro fuzzy inference system controller is 97.68%, 99.32% and 99.88% respectively, whereas the same without any controller is found to be 88.06.Item Design of Internal Model Control for Wastewater Treatment Plant Using Artificial Neural Network: Case Study Kality Wastewater Treatment Plant(Addis Ababa University, 2021-11) Haileyesus, Simegn; Dereje, Shiferaw (PhD)High non-linearity of wastewater treatment processes and the influent wastewater dynamics dependency on the season and surrounding lifestyle, it is still an active research area that requires a new control methodology to optimize and improve the current controller performance. Hence, this research is designed to develop an artificial neural network plant and controller models for the Kality wastewater treatment plant using feedforward backpropagation algorithms. Using the Levenberg Marquard learning algorithm, the plant model was developed to estimate the dissolved oxygen in the biological reactor with six input variables. After several ANN architectures are tested, the best result is obtained with three layers of 43 neurons each, a sigmoid activation function on the hidden layers, and a linear activation function in the output node. Similarly, an inverse ANN (internal model controller) model is designed to control the oxygen transfer coefficient with six input variables. The final optimal inverse neural network model structure is three layers with 47 neurons per layer, a sigmoid activation function on the hidden layers, and a linear activation function in the output node with the Levenberg Marquard learning algorithm. The simulation results also demonstrated that the performance of an ANN internal model controller outperforms that of a traditional PI controller in both transient and steady-state conditions. In the case of the PI controller, the plant response takes 19.4 seconds to reach a steady-state, whereas the ANN internal model controller requires only 16 seconds. This result shows that the ANN controller improves settling time by 21.25 percent. When the rise time of the plant response is evaluated, the ANN internal model controller improves the rise time by 67% compared with the PI controller. Since, ANN plant and controller models developed using process state variables, which provide a significant opportunity for disturbance rejection, transient and steady-state performance improvement.Item Fault Location Estimator Design for Power Distribution System Using Artificial Neural Networks(Addis Ababa University, 2018-07) Samuel, Shawul; Dereje, Shiferaw (PhD)Fault location in distribution system is critical issue to increase the availability of power supply by reducing the time of interruption for maintenance in electric utility companies. In this thesis fault location estimator for power distribution system using artificial neural network is developed for line to ground, line to line, line to line to ground and three phase to ground faults in distribution system. To develop this estimator one of rural radial power distribution feeder in Ethiopia, Oromia, Assela substation Gumguma line feeder is used as a test feeder. This feeder is simulated using ETAP software to generate data for different fault condition, with different fault resistance and loading conditions, which is the fault phase voltage and current. The generated data is preprocessed and put as an input for neural network to be trained. MATLAB R2016a neural network toolbox to train ANN and programming toolbox is used to develop graphic user interface for fault estimator. The feed forward multi layer network topologies of neural network with improved back propagation, Levenberg Marquardt learning algorithm is used to train the network. After the network (6-15-8-4) is trained the mean square error performance, regression plot and error histogram analysis was made and found to have an excellent performance with regression coefficient 0.99929 , validation performance of 0.000102 and error histogram range -0.015 to 0.019. In this thesis for practical implementation the fault records at the test feeder is handled by intelligent electronic device (IED) installed at the substation feeders. The fault record of IED can be read by PCM600 tool using laptop or manually using IEDs human machine interface, this fault recorded data feed to the graphic user interface to estimate the fault location as well as the fault type. Finally it is found that artificial neural networks are one of the alternate options in fault estimator design for distribution system where sufficient distribution network data are available with narrow fault location distance range from the substation. This has benefits in assisting for maintenance plan, saving efforts in fault location finding and economical benefits by reducing interruption time.Item Feedback Linearization with LQR Control Approach for Quadrotor Trajectory Tracking(Addis Ababa University, 2021-11) Betsega, Yosef; Lebsework, Negash (PhD); Dereje, Shiferaw (PhD)A quadrotor is a type of helicopter which is lifted and propelled by four rotor.Quadrotor’s are used for variety application due to reason of mechanically simple,small size and low inertia.But controlling quadrotor is a challenging task because of non linearity and underactuacted character of the dynamics. In this thesis work,the dynamic model of the quadrotor is derived,by using NewtonEuler formulation.Then a nonlinear control approach which is a feedback linearization technique is used . Feedback linearization technique is used to linearize both the translational and rotational dynamics.Here during linearization an auxiliary control inputs arises and those linear control inputs are found through linear control technique called linear quadratic regulator.Constant trajectory tracking and rectangular trajectory tracking is also obtained by controlling the attitude and the position simultaneously. The quadrotor dynamic model and the controller design is developed using MATLAB simulation and the dynamics model is tested with step input for each control inputs.The effectiveness of the proposed controller is tested with disturbance and without disturbance through MATLAB simulation.It is observed that,on the constant tracking simulation result the step input response have good settling time response.It achieve the desired 100m position on the settling time of x 19.42 sec, y 19.5 sec and z 11.4 sec .Item Fuzzy and PLL Based MPPT System Design for Energy Efficiency Of PV System:With A Case Study in Samara University(Addis Ababa University, 2020-01) Kiros, Abaye; Dereje, Shiferaw (PhD)Maximum power point tracking plays an important role for Photovoltaic systems because it optimize the power output from a Pholovoltaic system for a given set of conditions. This thesis presents a maximum power point tracker using artificial intelligence(Fuzzy controller) and Phase locked loop for a Grid Connected Photovoltaic system. The aim of this paper is to improve the dynamic performance of the available photovoltaic (PV) system and maximizing the power obtained from it by the use of cascaded converters with intelligent control techniques. Fuzzy logic based maximum power point technique have embedded in the first conversion stage to obtain the maximum power from the available PV array. The cascading of second converter which is needed to maintain the terminal voltage at grid potential. The soft-switching region of the three-stage converter is increased with the proposed phase-locked loop leads to a reduction in the ripple content, rating of components, and switching losses. The PV array is mathematically modeled and the results will be analyzed. This thesis lead to accomplishing maximum power and improved reliability for the same insolation of the PV system. After designed and sized the proposed system components, each System elements are individually modelled in MATLAB/SIMULINK and then connected to assess performance under different environmental conditions. The system consists of modeled PV Array, Boost converter model, TSSSBC, modeled fuzzy controller, Phase locked loop, inverter, and low pass filter is applied to generate a 24.22kW power by using a boost converter as a supporter for maximum power point traking algorithm by adjusting the duty cycle of the boost converter to maximize the output signal of the solar PV array was optimized using boost converter. The performance of the proposed controllers for energy efficiency of PV system is evaluated through simulation studies. The simulation results show that the efficiency of the photovoltaic system with the fuzzy logic controller is 99.2%, where as the same without any controller is found to be 87%.Item Model Predictive Control Design for Agricultural Robot Operation in Row Culture(Addis Ababa University, 2020-12) Asheberom, Gebreslassie; Dereje, Shiferaw (PhD)Agricultural robotic vehicles have the capacity to play a key role in the future of agriculture. For this to have controller designs that are cost effective and easy to use is very important. Agricultural robots which operate in row culture agricultural need to strictly follow the wheel tracks. The robot kind selected in this thesis is a differential drive wheel agricultural robot with one rear mounted caster and two front wheels. Navigation errors where the robot sways of its path with one or more wheels may damage the crops. Since model of the plant is paramount in designing of the MPC, mathematical model of the robot involves two identical series DC motors dynamics, the robot chassis dynamics and the robot kin-ematics. The model of the robot is simulated in MATLAB. The effect of change in mass of the robot is considered as disturbance for the better position and orientation straight crop row tracking. This paper focuses on the designing of MPC (Model Predictive Control) for agricultural robot operation in row cultures and then improves the performance of the robot to track its position and orientation from the desired crop line position. The two DC series motors found in each front wheels and the controller are simulated by MATLAB. The input component (pinhole camera) is specified. The robot position (x, y), orientation (α) and control signals (���� ,����) are constrained to restrict the forward speed and maximum error of the angle or heading. The performance of the proposed controllers for position and orientation set point tracking is evaluated through simulation studies. The simulation results show that the cost of tracking the desired position (����,����) and orientation (����) of the crop in the row is ��=1.0653∗��−10. The lowest cost function means the lowest error between the desired and actual position, orientation. The MPC approach is very advantageous and display better performance when facing the path constraints of operating in agricultural which follow row culture.Item Model Predictive Control of Unmanned Aerial Vehicle for Locust Detection and Bio-pesticide Spraying(Addis Ababa University, 2021-11) Eden, Getiye; Lebsewerk, Negash (PhD); Dereje, Shiferaw (PhD)Swarm of Locust are very harmful for food security, quality and quantity of agriculture products. Ethiopia is one of the countries which is extensively affected by locust invasion. The locust swarms have destroyed large swaths of food and pasture in Ethiopia which lead to famine and displacing thousands of people from their home. Ethiopia battled the swarms by spraying pesticides from air using helicopters leased from FAO. With this consideration, precise locust detection and bio-pesticide spraying is significant for preventing locust plagues. This thesis is going to focus on the design of Model Predictive Control of UAV for locust detection and bio-pesticide Spraying. To accomplish this design: First the dynamics of the system was understood then the mathematical model of the system was done and it was based on an agriculture spray drone (JMR-X1400). The Newton-Euler formalism was used to model the dynamic system and verified in Simulink. The flight controller is designed and MPC is implemented for this thesis. For this non-linear dynamic system of a quad-copter NMPC (non-linear MPC) is chosen. Multiple shooting method is selected to transform the optimal control problem to nonlinear program (NLP). To solve the NLP, CasADi in MATLAB is used and the solver is Ipopt (Interior Point Optimizer). The NMPC was able to control the quad-copter, which means the quad-copter was able to follow the given reference trajectory with minimum control effort. Since the quad-copter is used to spray pesticide, there will be a change in mass when it sprays. For this reason the Recursive Least Square Estimation (RLSE) is used to estimate the mass change and the model can be updated using the estimation. The proposed method works adequately. The RLSE was able to estimate the mass change and the quad-copter was still able to track the reference. Manual monitoring is a labor-intensive job and expensive for large farms. To tackle this problem, image recognition have provided a promising solution for detecting pests. So for this thesis Image recognition system is developed to detect and recognize the Locust swarm. Since it is an Image classification, CNN is chosen and the programming language is in python. After passing through different procedures the final training accuracy of the machine is 95:19%.Item Modeling, Design and Control of The Belt Drive System In Finchawa Sugar Factory(AAU, 2017-12) Deressa, Tesfaye; Dereje, Shiferaw (PhD)This thesis presents the modeling, design and control of the belt drive system in Finchawa sugar factory. Belt drives have been serving in this industry for a long period of time. Certain features of belt drives such as slippage, tension fluctuations, and sliding of the belt on the pulleys lead to highly nonlinear deformation, large rigid body motion, dynamical contact with sticking and slipping zones and cyclic tension. In addition to these problems, in Finchawa sugar factory, the tracking control mechanism of the belt drive system has many problems. The main problem is slow response and lack of accuracy. This is because the control is manual with simple ON/OFF control mechanism which is more difficult to obtain fast response and accurate tracking. The performance of motion control for belt drives is important in many industrial fields and is affected by these factors. PI control can improve performance specification of belt drive system and result in a faster dynamic response and more accuracy. Hence in this thesis, modeling of a linear belt-drive system and designing of PI controller for its position and speed control mechanisms has been examined by changing the reference of the system with the maximum steady state error of 0.0027% and good transient performance with rise time less than 0.1 second. Friction phenomena and position dependent elasticity of the belt was analyzed. The PI controller has been designed for accurate speed and position control mechanism and was simulated on MATLAB.Item Neural Network Based Direct Model Reference Adaptive Control Technique For Improving Tracking Performance in Nonlinear Systems.(Addis Ababa University, 2019-07) Alemie, Assefa; Dereje, Shiferaw (PhD)This thesis investigates the application of a neural network based model reference adaptive intelligent controller for controlling of the nonlinear systems. In this scheme, the intelligent supervisory loop is incorporated into the conventional model reference adaptive controller framework by utilizing an online growing multilayer back propagation neural network structure in parallel with it. The idea is to control the plant by minimizing the tracking error between the desired reference model and the nonlinear system using conventional model reference adaptive controller by estimating the adaptation law using a multilayer back propagation neural network. In the conventional model reference adaptive controller (MRAC) scheme, the controller is designed to realize the plant output converges to reference model output based on the plant, which is linear. This scheme is effective for controlling the linear plant with unknown parameters. However, using MRAC to control the nonlinear system in real time is difficult. The Neural Network is used to compensate the nonlinearity of the plant that is not taken into consideration in the conventional MRAC. The proposed neural network based model reference adaptive controller can significantly improve the system behavior and force the system to follow the reference model and minimize the error between the model and the plant output. Adaptive law using Lyapunov stability criteria for updating the controller parameters online has been formulated. The effectiveness of the proposed control scheme is verified by developing the simulation results for simple pendulum and Vander poll oscillation as a benchmark study in MATLAB/SIMULINK software. It is observed from the simulation results that the proposed neural network based Direct MRAC has 3.13sec rise time, 5.15sec settling time for 0.1rad disturbance and 3.12sec rise time, 5.21sec settling time for 0.2rad disturbance. Whereas, the conventional direct model reference adaptive control has 5.42sec rise time, 15.5sec settling time for 0.1rad disturbance and 5.01sec rise time, 15.52sec settling time for 0.2rad disturbance. It is shown that the proposed neural network based Direct MRAC has small rising time, steadystate errorItem Neural Network-based Smart Meter Demand Response Analysis: A Case Study of Addis Ababa Power System(Addis Ababa University, 2021-10) Betelhem, Abera; Dereje, Shiferaw (PhD)Most African countries including Ethiopia used the old way of going door to door to record usages of electricity. This results lots of guesswork, which has direct impact on consumers, especially in billing. In case of Ethiopian Electric Utility (EEU) getting real-time information about power interruption, maximum power consumption in the grid is difficult to maintain and implement especially when going down towards the end consumers. The concepts of Smart Meters are introduced to address problems associated in electricity transmission throughout the usual traditional grid, which allows a bidirectional communication between the household smart meters and the supplier. This study aims to explore demand response analysis of smart meters using available recorded information by training Neural Network method to identify maximum demand response, type of power interruption and identify theft, by means of Artificial Neural Networks (ANNs) with Feedforward backpropagation algorithm for the selected cases in Addis Ababa as a case study. Four districts in the EEU Addis Ababa City were used for collecting quantitative data. The result for theft identification purpose consists of 169,296 samples, 25 neurons, two outputs. The best validation performance is 0.003124, and the overall correctly predicted percentage becomes 99.7%. In power fluctuation classifications, the model data sets consist of 3,596 sample sizes with 30 hidden neurons. The best validation performance is 0.03197. Moreover, the overall percentage of correctly predicted values is 97.2%. Finally, for the maximum power demand the percentage of correctly predicted values is 100%. The data analysis highly affected the performance of the NN system. Lastly the study recommends, further improvements can be achieved by process real-time data from millions of smart metering, more efficient modeling can lead to higher prediction accuracy.Item Non-linear Adaptive Artificial Neural Networs Control of Municipal Wastewater Treatment Plants(Addis Ababa University, 2019-06) Solomon, Baye; Dereje, Shiferaw (PhD)Wastewater is used water from any combination of domestic, industrial, commercial or agricultural activities, surface runoff or stormwater, and any sewer inflow or sewer infiltration. The characteristics of wastewater vary depending on the source. Types of wastewater include domestic wastewater from households, municipal wastewater from communities (also called sewage) or industrial wastewater from industrial activities.Wastewater treatment is the process of treating contaminants prior to releasing wastewater into the environment or reusing. Basically, there are four steps to remove contaminants in sewage wastewater which are; pretreatment, primary treatment, secondary treatment and tertiary treatment.The activated sludge process is a biological process and an essential secondary treatment in wastewater treatment , where bacteria plays a role of degrading organic substances based on the the crucial process control parameter, dissolved oxygen (DO) concentration. The DO concentration in the aeration tank(s) is maintained at the desired level by manipulation of airflow rate, applying a Neural network based adaptive Proportional-Integral-Derivative (PID) controller. In this thesis work, an Adaptive Neural Network Radial Basis Function PID (ANNRBFPID) control strategy is implemented to control a DO concentration in aerated bioreactors which update the set point of DO adaptively and withstand uncertain disturbances. Two models are selected to represent an activated sludge process. The first one is the simplified model with only four state variables. The second model is the Activated Sludge Model no.1(ASM1) the more realistic and accepted model with 13 state variables. Matlab/Simulink and SIMBA# software used for simulating the designed mathematical model and control of the activated sludge process for the simplified model and ASM1 respectively. The powerful learning and adaptive ability of the RBF neural network make the adaptive adjustment of the PID parameters to be realized. Hence, when the wastewater quality and quantity fluctuate, adjustments to some parameters online can be made by ANNRBFPID algorithm to improve the performance of the controller. The Matlab/Simulink simulation result show that the DO can be maintained at 2mg/L or any desired setpoint with the presence of uncertain disturbances and continuously variable influents with ANNRBFPID control algorithm and the simulation result shows that ANNRBFPID achieve better control performance than conventional PID. On the other hand, SIMBA# simulation results show that the international standard limit for Ntot (Total Nitrogen), CODtot (Total Chemical Oxygen Demand), SNH (NH4(+) and NH3 nitrogen), TSS(Total Suspended Solids) is given by < 18g, < 100g, < 4g, < 30g respectively and the simulation result obtained is 11.04 g N/m³, 23.82 g COD/m³, 0.5421 g N/m³, 5.061 g/m³ respectively.