Vacuum Level Acquisition System of BP Neural Network Based on DSP

TIAN Jinbo; ZHAO Zhizhong; BI Haiyan

doi:10.13296/j.1001-1609.hva.2014.12.009

您当前的位置：

首页 >

文章列表页 >

Vacuum Level Acquisition System of BP Neural Network Based on DSP

更新时间：2025-04-26

- Vacuum Level Acquisition System of BP Neural Network Based on DSP
- High VoltageApparatus Vol. 50, Issue 12, Pages: 47-54(2014)
- 作者机构：
  
  1. 河北工业大学电磁场与电器可靠性省部共建重点实验室
  2. 天津市电力公司城东供电分公司
- 作者简介：
- 基金信息：
- DOI：10.13296/j.1001-1609.hva.2014.12.009
  CLC： TM561.5
- Published：2014
- 稿件说明：
移动端阅览
[1]田金波,赵智忠,毕海岩.基于DSP的BP神经网络真空度数据采集系统[J].高压电器,2014,50(12):47-54.

TIAN Jinbo, ZHAO Zhizhong, BI Haiyan. Vacuum Level Acquisition System of BP Neural Network Based on DSP[J]. High voltageapparatus, 2014, 50(12): 47-54.
[1]田金波,赵智忠,毕海岩.基于DSP的BP神经网络真空度数据采集系统[J].高压电器,2014,50(12):47-54. DOI： 10.13296/j.1001-1609.hva.2014.12.009.

TIAN Jinbo, ZHAO Zhizhong, BI Haiyan. Vacuum Level Acquisition System of BP Neural Network Based on DSP[J]. High voltageapparatus, 2014, 50(12): 47-54. DOI： 10.13296/j.1001-1609.hva.2014.12.009.

摘要

设计了基于DSP芯片TMS320F2812和BP神经网络数据融合的真空度信号数据采集融合系统。首先

利用横磁场脉冲磁控法和屏蔽罩电位法双传感器

对真空灭弧室进行真空度检测

采集的模拟信号经放大处理后通过DSP转换到数字信号;其次

利用BP神经网络方法进行数据融合

以获取准确的采样数据。BP神经网络的训练过程需要的时间长、数据量大

所以利用特征向量进行训练的环节在MATLAB上实现

将训练好后得到的权值矩阵和阈值向量嵌入到DSP中;然后

对真空灭弧室的真空度采集数据处理后的融合值由TMS320F2812的e CAN模块串行通信到上位机由VB可视化页面对数据进行检测分析。重点提出的基于DSP的BP神经网络数据融合技术的数据处理方法和串口通信

并实验证明这种方法具有良好的容错性

诊断结论的可信度明显提高

能够有效提高测量数据的稳定性和准确度。

Abstract

The system is designed in order to acquire and fuse the vacuum data that is based on DSP TMS320F2812 and BP neural network.First

in order to obtain accurate data

the pulsed magnetron discharge method and the improved potential of shielding case method are chosen.After the vacuum level of the vacuum interrupter is detected

the acquired analog signal which is enlarged is processed by DSP conversion to the digital signal;Secondly

with BP neural network the data is fused in order to obtain accurate sampling data.Training BP neural network process would take a long time and amount of data

so feature vectors is trained in MATLAB.The weight matrixes and thresholds vector are embedded in DSP;Last

after fusing

the higher vacuum interrupter fusion data are communicated by the TMS320F2812’s e CAN module then the data is communicated with PC and analyzed by the visual of VB page.The data fusion method of BP neural network fusion technology and serial communication are focused in this system.The experiment shows that this method has high fault tolerance and the reliability of diagnostic conclusions can be obviously improved.Also the stability and accuracy of measurement data can be effectively improved.

关键词

Keywords

references

Views

224

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Simulation Design and Optimization of Permanent-magnet-motor Operating Mechanism for Vacuum Circuit Breaker with Long Clearance

Study on Axial Magnetic Field Measurement and Magnetic Field Saturation Characteristics of Iron Type Bipolar AMF Contact

Electric Field Analysis and Calculation of Parallel Capacitors for 252 kV Environment-friendly GIS Double-break Vacuum Circuit Breaker

Research and Improvement on Resistance to Arc Ablation and Melt Welding of CuCr Contacts in Vacuum Interrupter Chamber

Fault Diagnosis Algorithm on Spring Operating Mechanism for High Voltage Circuit Breaker and Load Switch Based on Motor Current

Related Author

LI Jinqi

DING Yiling

WANG Zhoulin

DING Can

ZHAO Qi

WANG Jun

LIU Shaoyong

WANG Zihan

Related Institution

College of Electrical Engineering & New Energy, China Three Gorges University

State Key Laboratory of Electrical Insulation and Power Equipment, Xi'an Jiaotong University

China Yangtze Power Co., Ltd. Wudongde Hydropower Station

China XD Electric Co., Ltd.

Xi'an High Voltage Apparatus Research Institute Co., Ltd.

⁰