微型智能低空速传感器的设计

利用大气中的两个物理参量大气静压力、大气动压力和空速的关系,分别用两个高精度压阻式绝对压力传感器对大气静压力和大气动压力进行测量,再用温度传感器对两个绝对压力传感器的测量值进行温度修正,将修正值作为多层感知机神经网络的输入,空速值作为输出.对标准或实验数据采用动批量法进行网络训练,得到网络各层的权值和偏置值,由此可以确定在任意静压力和动压力输入作用下,网络输出估计的空速值,最后借助数字信号处理器芯片实现微型智能低空速传感器的设计.实验结果表明通过训练、学习的传感器具有更高的精确度,同时能够很好地抑制时间漂移.并且很好地解决了低空速测量误差大的问题.

The Design of A Novel Micro Intelligent Low Airspeed Sensor

This paper designs a novel micro-intelligent airspeed sensor. Based on the relationship between the airspeed and the two physical parameters such as the static air pressure and the total air pressure, the static air pressure and total air pressure are measured out by the high accuracy piezoresistive absolute pressure sensor respectively.Then,the measured values of the two sensors are modified by the temperature sensor. Finally, the modified values are specified as the inputs and the airspeed as the output of the multilayer-perceptron. The weights and the offsets of each layer are obtained after training the artificial neural network(ANN) by applying Momentum Backpropagation algorithem(MOBP) using the stardard data or the experiment data. As a result, the network could output estimated airspeed values under a given static and total air pressure , so the new micro-intelligent airspeed sensor is realized with the help of the digital signal processing chip(DSP).The experimental results indicate that the new sensor can provide a higher degree of accuracy in restraining sensors from temperature shifting and time shifting, moreover, it reduces essentially the error of the low airspeed.