Converter Design for Solar Powered Outdoor Mobile Robot Josue Cruz-Lambert, Patrick Benavidez, Jacqueline Ortiz, Jack Richey, Shane Morris, Nicolas Gallardo, and Mo Jamshidi Department of Electrical and Computer Engineering The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249, USA Email: josueilambert@yahoo.com, patrick.benavidez@utsa.edu, jaqdandori@gmail.com, xje791@my.utsa.edu, shane.morris33@yahoo.com, hbq744@my.utsa.edu, moj@wacong.org Abstract—This project presents a hybrid system implementing the use of solar panels and batteries to power a robot. The main aim is to integrate a charging system which allows the batteries to be charged from solar panels, wall outlet, and a deployable solar charging station. The proposed system is divided in three sections: design of solar panels, design of a battery charger and design of a DC-DC converter with fuzzy MPPT tracking system. This paper will cover only up to the design of the DC-DC converter, as further work is still pending design/results. For the design of the solar panels, different cell configurations were considered. Once the solar panels were designed and fitted to the robot, we determined the battery requirements to meet the robot power load and payload. Lithium Polymer batteries were chosen as the power source for the robot since they have a competitive power density to weight ratio. In order to extend the battery life, and simplify the load, we decided to take two battery banks. One battery is charging while the other battery is discharging. This allowed us to precisely control the battery charging profile without load variations interfering with our measurements. We used a fuzzy maximum power point tracking controller with a primary focus on regulating the power input for a lipo. Different topologies of DC-DC converters were considered and based on our literature research, it was concluded that a Buck-Boost converter is the most appropriate option when working with solar cells. The following paper discusses all the design specifications, component decisions and construction of the solar powered robot. Various technologies, not used in the robot, are included here as a literature review of the current state of the art. The goal of this paper is to summarize the tested methods and results to expedite future researchers in the correct direction. Index Terms—Solar, Renewable, LiPo, Lithium Polymer, Robotics, Fuzzy Controller, Energy I. I NTRODUCTION This paper is a summary of the various techniques and technologies used to design a solar powered robot. Although no final results are included, this paper explores the current available products, software and their limitations for this application. The aim of this paper is to bring new college up to date with the current state of our research.up to date with my research, in turn saving them the time involved with the discovery process. Section two will explain the goals and objectives of the project. The following sections are organized in the chronological order that they were explored in. II. DESIGN CONSTRAINTS In this project, we are interested in installing a solar panel on a mobile robot platform. The main objective is to have the robot operate outdoors without the need of external power supplies. To start off the debate, we wanted a robot that would operate 100% on solar power. However, the solar cells an individual can get a hold off are only up to 15% efficiency. Attempts to acquire solar cell from reputable manufacturers went unanswered. Since our initial focus was to acquire the most efficient cells for their weight, we had to refocus our goal on what would actually fit on the robot. Since we were limited by the physical space of the robot, we considered optimizing the structure of the cells. Some form factors are pyramid arrangement, inverted pyramid, accordion, and flexible membranes. After ordering and producing a small sample of two cell configurations, we determine that a flat solar panel would be the easier to manufacture, marginally more efficient than the other configurations. Once we decided on the solar cell structure, we began to construct a solar cell panel. As a forewarning to the reader, construction of the solar cells is a time consuming endeavor, approximately 28 hours per panel. Building a solar cell on the sole merit of saving money is not a good reason to make your own after factoring in the time, cost, and materials. Our recommendation would be to spend more on an off-the-shelf component if possible. From the initial batch of solar cell, we ordered 100 units to create our panel. In the selection process, we picked the solar cell that claimed to have the best watts/meter rating. Normalizing the results, and then selecting the best for our robot. After receiving them, we started to characterized them, and record their performance. More details are provided in section four, solar panel modeling. Our next focus was the load the robot would take, having a fixed power generation, we determine the best approach was to maximize our power usage. To simplify the equations, we restricted the solar panel to exclusively charge lithium-polymer batteries. This would in turn reduce the complexity of the circuit to two components, the buck-boost converter, and a single battery to charge. To further reduce the load disturbance on a panel, we added an additional bank of batteries. The goal is to have one bank always charging, therefore only having the