Artificial Neural Network based Battery Management System on State of Charge Estimation for Optimal Operation of Photovoltaic-Battery Integrated System
Due to the reduction of fossil-fuel utilization PV-Battery integrated system is a preferable power supply in many areas of the world. Designing a supervisory controller that can harvest high energy density and prolong the battery lifetime is one of the major challenges in a battery energy storage sy...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/33252/4/Artificial%20Neural%20Network%20based%20Battery%20Management%20System%20on%20State%20of%20Charge.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Due to the reduction of fossil-fuel utilization PV-Battery integrated system is a preferable power supply in many areas of the world. Designing a supervisory controller that can harvest high energy density and prolong the battery lifetime is one of the major challenges in a battery energy storage system. A proper Battery Management System (BMS) monitors the battery charge status and takes decision to lengthen the battery lifetime. A regulatory State of Charge (SOC) estimation based on PV-Battery standalone system is presented in this research that significantly addresses the issues. The proposed control algorithm estimates SOC by Backpropagation Neural Network (BPNN) scheme and implements Maximum Power Point Tracking (MPPT) system of the solar panels to take decision for charging, discharging or islanding mode of the Lead-Acid battery bank. The proposed model is designed in MATLAB/SIMULINK software and the experimental prototype is assessed via dSPACE 1104 component. The proposed power control strategy is explored as robust as well as attained the effective objective of standalone PV-Battery Management System e.g. avoiding overcharging and deep-discharging manoeuvre under different solar radiations and temperatures. A case study is presented for several SOC estimation methodologies that demonstrate the effectiveness of the proposed strategy with 0.082% error. |
|---|