摘要
在车载式振动压路机压实度检测中利用加速度传感器对信号的检测,由于加速度传感器受自身性能、安装位置、方向、土壤压实度和振动压路机振动轮振动量等随机性、不确定性和模糊性的环境因素的影响,使得加速度传感器的信号输出是一种典型的非线性系统。RBF神经网络因具有较强的自组织性、自学习能力和自适应性等优势,更适合对加速度传感器的输出进行仿真与预测。基于Matlab程序建立了加速度传感器的神经网络模型,通过神经网络的优势性能对加速度传感器在信号检测中进行预测。
Detect signal with the acceleration sensor in the testing of Vehicle-mounted vibratory roller compaction. The output of acceleration sensor is a typical nonlinear system, because acceleration sensor is influenced by its performance ,installation position, installation direction,soil compaction and vibration roller wheel＇ s vibration frequency. And these factors are largely random. RBF neural network has strongly automatic arrangement,automatic learn ability, and automatic adaptive function, so it can be good at emulating and predicting the output of acceleration sensor. This paper established acceleration sensor neural network model based on the Matlab program. Predict the output of acceleration sensor with neural network in the detection.
出处
《仪器仪表用户》
2012年第2期32-34,共3页
Instrumentation
