基于BP神经网络的超声测距误差补偿  

Ultrasonic distance measuring error compensation based on BP neural network

作　　者：李飞[1] 陈峰[2] 

机构地区：[1]山东工商学院教务处计算机中心,山东烟台264005 [2]辽宁工程技术大学电气工程系,辽宁阜新123000

出　　处：《传感器与微系统》2006年第6期31-33,共3页Transducer and Microsystem Technologies

摘　　要：指出了超声波在测距应用中的局限性,并给出解决方案。着重从新的角度补偿超声传感器的误差,提出了用BP前馈神经网络补偿超声波声速受温度、湿度变化而引起的误差。在室外工作的测距仪上,应用该补偿方法后超声测距的精度提高了2个数量级。本方案适用于高精度要求或复杂环境下超声测距。The limitation of the ultrasonic sensor in distance measuring is analyzed ,and the scheme solved is put forward. The ultrasonic velocity sound error caused by change of temperature and humidity is compensated by BP neural network. Applied this method to ultrasonic distance measuring in the outdoor, the precision is up two grades. The proposal is suitable for ultrasonic sensor at complex environment or high precision.

关 键 词：BP神经网络 超声测距 补偿 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]