基于振动声学的除焦状态检测系统
|
摘要
针对传统水力除焦操作中需要人为观测判断、劳动强度过大、环境恶劣的缺陷,提出了基于振动声学的智能除焦状态检测系统,利用智能检测技术完成除焦状态的判读。通过振动传感器对信号进行采集,再完成特性参数的提取,使用模式识别的方式建立除焦状态和振型参数之间的关系,采用BP神经网络将获取的振动信号样本进行傅里叶变换后得到振动信号的幅频曲线。提取不同特征频段的幅值作为特征参数,进行样本学习训练,建立起除焦状态与焦炭塔振动特性之间的学习识别网络,经过训练并且稳定的网络即可用于水力除焦智能检测系统中。
Aiming at the defects of traditional hydraulic decoking operation,such as human observation and judgment,excessive labor intensity and harsh environment,a vibrational acoustics-based intelligent decoking state detection system is proposed,which uses intelligent detection technology to complete the decoking state interpretation.The signal is collected by the vibration sensor,and the characteristic parameters are extracted.The relationship between the decoking state and the vibration mode parameter is established by pattern recognition.The amplitude-frequency curve of the vibration signal is obtained by applying BP neural network to Fourier transform of vibration signal sample.The amplitude of different characteristic frequency bands is extracted as the characteristic parameters,and the sample learning training is carried out to establish a learning recognition network between the decoking state and the vibration characteristics of the coke tower.The trained and stable network can be used in the hydraulic decoking intelligent detection system.
Aiming at the defects of traditional hydraulic decoking operation,such as human observation and judgment,excessive labor intensity and harsh environment,a vibrational acoustics-based intelligent decoking state detection system is proposed,which uses intelligent detection technology to complete the decoking state interpretation.The signal is collected by the vibration sensor,and the characteristic parameters are extracted.The relationship between the decoking state and the vibration mode parameter is established by pattern recognition.The amplitude-frequency curve of the vibration signal is obtained by applying BP neural network to Fourier transform of vibration signal sample.The amplitude of different characteristic frequency bands is extracted as the characteristic parameters,and the sample learning training is carried out to establish a learning recognition network between the decoking state and the vibration characteristics of the coke tower.The trained and stable network can be used in the hydraulic decoking intelligent detection system.
引文
[1]王涛,高志,叶健敏,等.水力除焦监测系统的设计和研究[J].工程设计学报,2010,17(2):146-150.
[2]林广田,晋西润,解学仕,等.延迟焦化装置水力除焦系统自动控制技术[J].炼油技术与工程,2006(12):47-50.
[3]马雪勇,陈建新.延迟焦化智能除焦系统的研究与应用[J].知识经济,2013(11):110-111.
[4]刘霞,王利恒.除焦状态智能检测技术研究[J].自动化与仪表,2016,31(3):5-8.
[5]沈花玉,王兆霞,高成耀,等.BP神经网络隐含层单元数的确定[J].天津理工大学学报,2008(5):13-15.
[6]瞿国华.延迟焦化工艺与工程[M].北京:中国石化出版社,2008.
[7]王荣基,李爱兰.延迟焦化装置联合水力除焦控制系统[J].炼油化工自动化,1997(2):24-29,53.
[2]林广田,晋西润,解学仕,等.延迟焦化装置水力除焦系统自动控制技术[J].炼油技术与工程,2006(12):47-50.
[3]马雪勇,陈建新.延迟焦化智能除焦系统的研究与应用[J].知识经济,2013(11):110-111.
[4]刘霞,王利恒.除焦状态智能检测技术研究[J].自动化与仪表,2016,31(3):5-8.
[5]沈花玉,王兆霞,高成耀,等.BP神经网络隐含层单元数的确定[J].天津理工大学学报,2008(5):13-15.
[6]瞿国华.延迟焦化工艺与工程[M].北京:中国石化出版社,2008.
[7]王荣基,李爱兰.延迟焦化装置联合水力除焦控制系统[J].炼油化工自动化,1997(2):24-29,53.