Calculation Of Capacitance For Ni-MH Battery Based On Fuzzy Neural Network And Research Of Battery Management System

To resolve the air pollution caused by automobile and energy safety, the United States, Japan and European Federation have established their own long-term program to develop the hybrid electric vehicles(HEV). The same research is done in China.Our government has put a lot of money into the research of HEV in"863"item.Battery provides accessorial energy for HEV. So the precise measure of state of charge(SOC), prediction malfunction of battery alarm as well as communication with HCU are all very important for the research of HEV. So It is important between theory and practice to do the research in this paper.To satisfy the requirement of practice, We have finished some aspects,follow as:①Introducing some electrochemistry concepts of Ni-MH battery, and do a pertinent experiment for Ni-MH battery under practical working condition of HEV. Based on aboved work, analyzing experiment datum of Ni-MH battery and summarizing some conclusions on performance of Ni-MH battery.②Building up the model of the Fuzzy Neural Network, using experiment datum of Ni-MH battery to amend the fuzzy control rules and optimize the membership function. validating dynamic SOC under different discharge condition of Ni-MH battery by using variational electric current.③To satisfy the requirement of Battery Management System(BMS), working out electrocircuit drawing and PCB drawing. working out hardware of BMS based on PIC18F458 and checking out and debugging electrical connection.④finishing software design of BMS, it contains some aspects,follow as: 2 secs. delay, pertinent parameters test, CAN and serial communication, arithmetic of SOC as well as main program. Based on aboved work,'MPLAB'software for PIC microcomputer is used to debug program of the collection electrocircuit subsystem and the host control electrocircuit subsystem,'watch'function are used to observe the value of pertinent registers and some mistakes are amended.⑤Building up experiment bench of BMS test, checking out some pertinent functions of BMS. Such as collection function of battery parameter of collection electrocircuit system, the CAN communication between collection electrocircuit system and the host electrocircuit system, serial communication of the host electrocircuit system and so on.Finally,Using variational electric current method to check the model of the Fuzzy Neural Network in this paper. The error between the calculational SOC and real SOC satisfy requirement. Doing pertinent experiment for BMS, it can send experiment datum of battery into PC successfully and The author in this paper has applied a patent for the BMS.