DESIGN AND IMPLEMENTATION OF THERMAL MANAGEMENT SYSTEM OF LiFeMnPO4 BATTERY WITH OPTIMIZATION USING SUPPORT VECTOR MACHINE

• Main Author: HUGUN SAPUTRA - NIM: 23315309, LINO
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/22910
• Institution: Institut Teknologi Bandung

The electric vehicle is an appropriate projection for use as a future transportation. This is due to the many advantages provided by this type of vehicle than a vehicle with a hydrocarbon combustion engine. Among them are the higher energy efficiency, environmentally friendly, reduce noise, and reduce dependence on fossil fuels. Meanwhile, lithium-ion batteries is the best candidate in its use as an energy provider on the electric vehicle. Some of the advantages of this type of energy storage are high energy density, no memory effect, and high cycle life. <br /> <br /> Although there are many advantages of the use of electric vehicles, but there are still many challenges to be solved. Such as battery price, distance per charge, reliability, safety, capacity, cycle life and charging time. So the performance of electric vehicles depends heavily on the battery. Some of the disadvantages of lithium-ion batteries are sensitive to excessive heat, unbalance cells, over discharge, and over charge. <br /> <br /> Basically the problem of excessive heat on lithium-ion batteries sourced on three things: the heat from chemical reaction inside the battery, the heat caused by following ohmic law, and the heat caused by ambient temperature. If excessive heat appear on batteries, but there is no attempt to remove it, it will cause a reduction in battery life or State-of-Health (SoH) of the battery, reduction in efficiency of the battery, or it will cause thermal runaway because separator inside the battery melt and short circuit happen. Therefore, this study focuses on building a thermal management system of lithium-ion batteries with purpose that the battery can be maintained at a temperature between 25 - 28oC. <br /> <br /> This study aims to obtain a thermal management system in the lithium-ion battery module. The support vector machine (SVM) serves to provide an assessment of the use of parasitic energy, the average battery temperature after the cooling system is applied, and the difference in battery cell temperature during the discharge process. The support vector machine is also conditioned to control the fan speed during the discharge process so that the battery condition is maintained at 26,5°C. The support vector machine (SVM) is a method of classification and regression where the method works based on learning data whose purpose is to predict new data or data to be calculated. The SVM method is divided into two parts which are support vector classification (SVC) for problems requiring classification techniques and support vector regression (SVR) used in the case of regression. In order for the results of the SVM method to be optimal, some adjustments to the SVM parameters are required. These parameters include SVM type, kernel type, gamma (&#947;), epsilon (&#949;), and nu (n). <br /> <br /> The result from this study can be concluded that the use of thermal management system on lithium ion battery affects the amount of energy that can be stored inside the battery. This is evidenced by the amount of energy that can be drawn from the battery in the C5 discharge rate. When the thermal management system is not used the amount of energy that can be taken is 2895 kJ and when the thermal management system used shows a higher value of 3044 kJ. In addition, the use of support vector machine (SVM) for the case of battery thermal management system is nu-svc type, where the kernel type used is radial base function (RBF), with gamma (&#947;) = 0,33, epsilon (&#949; ) = 0,01, and nu (n) = 0,01. From the parameter values, it is known that SVM method gives an assessment that fan speed variation is best in 20% and 40% position. Fan position of 20% shows parasitic energy = 0,39%, battery temperature difference = 4,20oC, and battery average temperature = 26,10oC. While the fan position of 40% shows the value of parasitic energy = 0,64%, battery temperature difference = 3,90oC, and battery average temperature = 25,68oC. This research also obtained SVM parameters that can also be used as feedback on the control system in order to maintain the battery temperature at 26.5 ° C as the nu-svc type, where the kernel type is the radial base function (RBF), with gamma (&#947;) = 0 , 20, epsilon (&#949;) = 0,01, and nu (n) = 0,01. <br />