Design And Research Of Battery Management System For Drones

With the rapid and high-quality growth of national economy,the integration of traditional industries and new technologies has been further deepened.Drones as a new technology leads the industry to upgrade and change in many traditional industries(such as national defense,film television and logistics,etc.).Battery pack is the only energy source of drones.The reliability and safety of battery pack are important guarantees for well of drones.Battery Management System is the key equipment to ensure the operation of battery pack on stable and safe.It has gone through three stages:discrete,analog and digital-analog integration.In this paper,Battery Management System for drones is designed that based on digital-analog hybrid integrated circuit,and it is described in detail that engineering design scheme and implementation method.Several methods for estimating State of Charge were analyzed,contrasted and researched.Afterwards Support Vector Machine was used to estimate State of Charge.The performance of this estimation method was evaluated by using the battery data set published by NASA.Firstly,the function and performance index have been determined on Battery Management System of drones,and the hardware selection,algorithm evaluation and overall scheme design have been completed.Secondly,the hardware and software design of the Battery Management System is carried out.In the aspect of hardware design,the hardware circuit design and debugging on Front-End Analog Module,Main Control Module and Wireless Communication Module have been completed.In the aspect of software design,the software design and debugging of Driver Layer,Middleware Layer and Application Layer have been completed.Finally,the typical conditions for Battery Set Testing of Miniature drones have been designed,which combined with the test standard of USABC Electric Vehicle Battery Test Manual and the flying condition characteristics of drones.The actual test platform is built to complete the simulation work condition experiments,and the estimation of State of Charge has been completed under different work conditions by using Support Vector Machine.The actual system test results show that the system runs well and meets the requirements of the design scheme.