Thesis Title: Development and Evaluation of the System for the Greenhouse Water using Capacitive Deionization
Water and energy are the main factors in the socio-economic development of societies. With the increase in world population to more than 7 billion, the demand for food and consequently water is more than before. It is very important to find a sustainable method for supplying the required water that is economically and environmentally acceptable. In this research, for the first time, the process of capacitive deionization has been mathematically modeled and the best conditions for producing healthy water with the lowest cost to be used in different sectors are provided.
Thesis Title: Construction, Evaluation of stability and thermal performance of a Salt-Gradient Solar Pond Launched by Greenhouse Drainage
Solar energy is the largest resource of clean energy in the world and it is available in most areas. This energy is a durable energy that needs to be collected for use. An example for solar collectors is the solar pond that collect and store solar energy by high concentrated brine.Generally, solar ponds are divided into ponds with heat transfer and ponds without heat transfer. In salt gradient solar pools (SGSP), the saltwater doesn’t move upwards and mixes by heating, increasing the pool’s temperature. SGSP consists of three different regions. The upper-convective zone (UCZ), the non-convective zone (UCZ), and the lower-convective zone (LCZ). The LCZ layer operates as a heat source that we can use the heat stored in this layer.
Due to the consumption of fertilizers and not complete absorption by the plant, greenhouse drainage includes poisons and nutrients and a high TDS level. Mismanagement of greenhouse drainage water, its release into the surrounding environment, and its storage in wastewater ponds have created severe challenges for greenhouse owners and will cause many environmental problems.
In this research, a laboratory salt gradient pond was constructed with a square cross section (40*50) and 60cm deep. The pond was filled with ammonium nitrate and potassium chloride two types of the most common fertilizers in the greenhouse. Experiments will be conducted in the laboratory under the incident radiation from two halogen lamps acting as a solar simulator. Behavior of this pond was studied for investigation of the temperature and concentration profiles as a function of depth.
Thesis Title: Design and Manufacture of the Gripper of the Saffron Selective Harvesting Robot
Saffron is one of the most important products whose cultivation has a long history in Iran. So that many researchers have attributed the main origin of this valuable plant to our country. Saffron production has been common in Iran for a long time and it has been mentioned as red gold. Although most of the production is related to Iran, but due to the problems related to packaging and global marketing, countries such as Spain and even the United Arab Emirates buy this product in bulk, package it and sell it in the world market.
According to the above and maintaining the position of Iranian saffron in global markets, in this project, a special gripper of the saffron selective harvesting robot is proposed to maintain the saffron quality during harvest, harvest speed increases, harvest time decreases and decreases the labor force.
In general, the goals and necessity of designing and manufacturing of the gripper of the saffron selective harvesting robot can be summarized as follows:
Harvest saffron at the best time with the highest quality
Do not contaminate saffron flowers with dust during harvest
Eliminate manual harvesting and prevent the reduction of saffron quality in hand contact during harvesting
Prevent the loss of agricultural inputs by identifying the best harvest time
Maintaining the position of Iranian saffron in world markets
Thesis Title: Design And Build Control System For A Field Flying Robot
Nowadays, precision farming for producing high-quality products and proper usage of agricultural inputs have become important due to the development of agricultural mechanization and automation smart systems. The use of GIS and GPS systems in precision farming has led to develop the application of flying robots for obtaining low-cost aerial photos from farmlands and orchards in comparison with airplanes and satellites. In this study, an eight-paddle vertical flight octocopter flying robot is designed and constructed with X8 configuration and its control system has been studied. The construction principles of this type of robots are similar to quadrotor, except that there is an inversely attached engine under each engine which improves the ability of takeoff and stability of the robot in the air. In this research, PID controller is utilized to control the robot motion. Firstly, the mechanical, electronic, and control components of the robot primarily designed and drawn in SolidWorks Software. Then, the ADAMS and MATLAB were used together as software simulation in the loop for simulation and control of its flight. After that, according to the flight data which was transferred from the robot to the ground station by a Bluetooth module, the performance of designed octorotor in hover flight was evaluated in real situation. The robot’s performance in hover flight was of importance because suitable and proper aerial photographs taken from lands and gardens which can be used in precision farming operations could be achieved only in hover flight. Experimental results of this study illustrate acceptable performance of the robot in modeling and hover flight.
Thesis Title: Design, Manufacturing and Evaluation of A Field Flying Robot
In recent years, Vertical Take-off and Landing (VTOL) multirotor has attracted the attention of many researchers. Using a robot in a precision farming process is rapidly increased, since these robots have the lower cost in compared to the systems such as airplanes and satellites which can provide the required images for local farms and gardens. In this study, design, manufacture, and evaluate of a flying robot with eight rotors (Octorotor) for the agriculture usage is considered. To this end, dynamic modeling of the Octorotor is done and then, this model is simulated and verified using the ADAMS and MATLAB programs. Next, various ideas are examined and so, a main body which is fitted to the considered purpose is designed. The aerodynamic and mechanic simulations of this robot are performed in the SolidWorks software environment. The simulation results show the highest stress and strain which are happened in the motors’ seat. It should be noted that the body’s frame with the lower drag force is suitable for the out-door environment usage. After that, the robot is controlled by the 4.2 GHz radio control system whereas an auto-navigation facility is also considered for the robot. To determine the situation and position of the flying robot, online flight information is sent to the ground station by the 433 MHz transmitter and receiver. The robot camera records the images which are transmitted to the ground station by the 8.5 GHz image transmitter. To control the robot, a PID controller is used which their coefficients are manually adjusted. Finally, simulation and implementation results are evaluated in a hover flight condition. These results show that the designed Octorotor has a good performance in the considered flight maneuver.
Thesis Title:Design, Manufacturing and Control of a Swing Pruning Cutter along Power Lines
From past to present, trees are essential elements in the construction of green space since they provide many ecosystem services. But, some problems are created when they are placed right-of-way the power lines. These problems are caused due to contact tree branches with wires of power towers. This contact is one of the major reasons of interrupting power services. In addition, as a result of this contact, trees are also at risk. So, to avoid this, these branches need to be pruned regularly. To this end, the safe environment is the first prerequisite since this task is dangerous for many reasons. So, an assumed robotic system can be useful to do this task. In fact, the accuracy and quickness of operation is obtained using this system. Therfore, the use of this system is proposed. In this research, design, manufacturing and control of a swing pruning cutter along power lines were studied. To reach this goal, a chain saw system is designed in SOLIDWORKS software, which is carried by an Unmanned Aerial Vehicle (UAV). The assumed flying robot is an Octorotor that was manufactured in AGRINS laboratory of Tehran University. To control this system, a usual Proportional-Integral-Derivative (PID) controller is used. To perform simulations and control of the flying robot with suspended saw, ADAMS and MATLAB software programs are linked together. Then implementation was carried out in real conditions and the obtained results of simulation were discussed. The results were shown that the PID control system has been able to stable the flying robot with the added suspended chain saw.
Thesis Title: Design and Analysis of a Laboratory Model of Combined Vertical Wind Turbine for Using in Agricultural
Usage of energy in the world is growing rapidly and by consideration of the fact that main portion of this energy provided by fossil resources, the pollution increased more than ever. Today according to global heat crisis and also environment pollution the global effort for finding efficient alternative ways increased. Usage of renewable energies such as wind, sun, biodiesel, are more become common place. Iran has good potential of producing renewable energies and must be stand in same way with other countries to be able to provide energy for the future. In this research the main goal is using renewable energy for pumping water in agricultural use. For this reason vertical wind turbine in a small scale has been designed. H-rotor darrieus because of good efficiency has been chosen and for covering of starting weakness the savonius rotor added to provide starting torque. Design and simulation of both turbines done in Ansys and Q-Blade software. Finally combined turbine has been manufactured and compared within simulation and a good agreement has been found.
Thesis Title: Design and Manufacture of a Laboratory Prototype of Water Canals Dredger Robot
Water is an essential life-sustaining element. Considering the importance of fresh water for life, It should be avoided its waste as much as possible. Given the reduced natural water availability in semi-arid areas, complex water supply systems were developed. Irrigation channels are artificial linear structures in the landscape that transport water. One of the important management measures in relation to water channels is to control the removal of vegetation in the channels to prevent water loss in this area. The growth of plants within the channels will reduce the effective capacity of the channels for water transfer as well as the velocity of water flow inside the channels. As a result, water evaporates from within the channels, followed by water loss. Aquatic plant coverage are managed using chemical, biological, ecological and physical methods. The mechanical-mechanical control method, which can be done either manually or hard-working, or by a mechanical machine, such as a mechanical excavator, which damages the channel characteristics, causing damage to the walls of the channels and the cost of repairs. Another method for controlling aquatic plants is dredging and deepening of their beds. Dredging will save water and will prevent it from wasting. In this project, design and manufacture of a laboratory prototype of water canals dredger robot will be studied. To achieve this goal, the robot will be designed in software. Robot simulations will be performed according to real conditions. Finally, implementation will be carried out in real-world conditions and the results of simulation and reality will be examined.
Thesis Title: Design and Analysis of a solar Linear Move Irrigation System.
Solar Energy is the most popular renewable energy, which is available unlimitedly all around the globe. In many modern engineering devices, it is highly appreciated to harness the green solar energy, both since it is clean compared to fossil fuels, and since it is always accessible with no need to any external power line, specifically in field operations. Linear move irrigation systems continually need access to energy, while working in wide fields, for motion and water pumping. This paper presents design and analysis of a linear move irrigation system, which is fed by solar panels. The design of structure is introduced, the system dynamics is modeled by computer and the proper motor and solar energy systems are designed based on the model. Finally, the obtained results are discussed.
Thesis Title: Design, Implementation and Evaluation of a Film Processing-Based System for Detection,Tracking and Grading of Cucumber Fruits
Image processing today is the best tool for extracting features and analyzing position and ultimately making the right decision. Similarly, in the case of humans, information is transmitted through the eye to the brain, and the brain processes the final decision by issuing a decree and issuing a decree. The purpose of image processing is to implement the human mind’s performance against data and perform certain processes to extract the required attribute to achieve the preset goal. In this scheme, cucumber grading was considered. Grading criteria are: length, diameter, curvature, uniformity of diameter, whiteness, uniformity of color, pitting and cutting (fracture). Given the ability to recognize these criteria at low cost by image processing, as a non-destructive inspection method and that image processing creates a consistent quality in terms of ranking, this method was used as the basis for the grading system. In this paper, for 50 cucumber samples, manually all its geometric dimensions were calculated manually and then simulated and measured by image processing and the results were compared with an acceptable error of 1.5 to 3 percent.
Thesis Title: Dynamic Modelling and Control a Cleaning Robot Solar Panels
Today, the use of renewable energies has been welcomed by the public due to the danger of fossil fuels running out. Solar energy is one of the most important renewable energy sources. Since solar panels are placed at natural environments, there is always a problem of dust accumulation on the panels, which results in absorbed energy reduction. For cleaning of the panels, robotic cleaners can be used. For relocation of such cleaning robot between solar panel rows, an automated carrier mechanism is required. Hence, a robotic system for cleaner displacement is introduced, and its dynamics modeling and control are presented. To this end, the kinematics characteristics of the robot are introduced. Then, the dynamics model is derived based on energy methods. For model validation, the robot is also modeled and simulated in standard and verified software, ADAMS, and the results are compared. To control the robot in order to move between rows, position and trajectory controllers are introduced. For position control, a PID controller is designed. For trajectory control, first, a CTM Computed Torque Method) controller is designed, based on the derived dynamics model.Then, for less computational efforts, a TJ (Transposed Jacobian controller) is designed, using the kinematics model. Finally, to increase control performance, an MTJ (Modified TJ) controller is also designed. Finally, for verification and analysis, the controllers are simulated, using co-simulation of MATLAB Simulink and ADAMS model. The simulation results show satisfactory performance of the controllers, and can be used for further design analysis of the prototype.
Thesis Title: Design and Manufacturing of an Aerial Pollinator Robot for Walnut Trees
Sufficient pollination is one of the most important factors in fruit production. Insufficient natural pollination in the beginning years of establishment of a walnut orchard is the cause of using artificial pollination as a complementary method to overcome this problem. Walnut is a monoecious species in which male and female flowers are present on the same tree.Walnuts are a species of trees pollinated by wind. Lack of sufficient pollen in the early years of establishment of an orchard and no overlap between pollen shedding and stigma receptivity has led to the use of artificial pollination, particularly in the walnut orchards that they use tissue culture plants. Nowadays, drones are widely used, and can also be used in agriculture. In this research, dynamic modeling, fabrication, simulation and control of a quadrotor prototype for walnut artificial pollination are studied. This system consists of four motors and is called quadrotor. Design steps are performed in SOLIDWORKS and then transferred to ADAMS and ANSYS FLUENT. A PID controller is used to control this system, and its gains were determined by try and error. Then, for control design and performance analysis, MATLAB and ADAMS co_simulation was performed. The system was modeled in Ansys Fluent to perform computational fluid dynamics (CFD) simulation to predict pollen motion streams. According to the results of the simulations and controller system, the robot was exprimented in a walnut orchard. The pollens were collected manually and then mixed in different percentages with talc powder and poured into the tank. Three treatments of 50%, 60% and 80% of pollen were sprayed on three rows of trees by the robot. After analysis of the performance of the machine and comparing with the control treatment performed by backpack systems and operator, the results showed that the robot has a good performance for pollination of walnut trees.
Thesis Title: Estimation of Chlorophyll Content of Tomato Plant Leaves based on Deep Learning Technique with on-site Imaging
Nitrogen is one of the most consumed and vital elements in plant nutrition and its deficiency is seen more in plants than other elements. In general, 1-5% of plant tissues are composed of nitrogen. Most of the nitrogen required by plants is absorbed as nitrate, which plays an important role in the production of chlorophyll molecules and various plant proteins. The content of green plant pigment (chlorophyll) is an indicator of photosynthetic activity, which is one of the most basic biological activities of the plant and is closely related to the concentration of available nitrogen. By saving nitrogen consumption, in addition to helping the farm economy, environmental crises such as groundwater and surface water pollution are prevented. to plan the optimal use of nitrogen fertilizers, it seems necessary to have the chlorophyll content of the leaves. In this research, a deep learning method is proposed to estimate the chlorophyll content of tomato leaves at the plant site in a non-destructive way by imaging the plant leaves. This method consists of three main parts: deep convolutional autoencoder neural network for noise removal, deep autoencoder neural network for feature extraction, and support vector regression network for estimating leaf chlorophyll content. The main advantages of this work are non-destructive, user-friendly, and high accuracy for practical purposes. Noises in this system were automatically detected and removed using a deep cannulation self-coding neural network without manual intervention. The amount of loss during the training and validation stages of this network converged to 0.5021 and 0.5317. Also, the evaluation of this network based on the binary cross-entropy criterion was significantly reduced and converged to 0.55. To select the best features from the leaf image, a deep self-coding neural network was designed and used to establish the best fit between the image of each leaf and its measured chlorophyll number. The network automatically learns and extracts the best features from each image without having to change the feature extraction algorithm. Loss values during training and validation stages with mean square error (MSE) converged to 0.0251 and 0.0292, respectively. The amount of loss in the evaluation phase of this model was converged to 0.45 based on the binary cross-entropy method. The features extracted from the image of each leaf along with the corresponding chlorophyll values were provided as input to the backup vector regression network to teach and learn the relationship between image features and chlorophyll content. The loss value of this model converged to 0.087 with an average square error using the K-fold cross-validation method with a value of k equal to 4.
Thesis Title: Design and Manufacturing of the Ink and Injection System for a Chocolate 3D Printer