湿度传感器的SVM温度补偿研究及软件设计

Software Design and Study of Humidity Sensor SVM Temperature Compensation

针对湿度传感器受温度影响严重而导致测量精度下降的问题,提出了采用改进的粒子群(PSO)和支持向量机(SVM)相结合的方法(AMPSO-SVM)对传感器进行温度补偿,建立了补偿模型,并与传统的网格寻优支持向量机(GRID-SVM)以及遗传算法支持向量机方法(GA-SVM)进行了比较。结果表明:改进的粒子群支持向量机方法能有效地降低温度影响,提高了湿度传感器的测量精度,并且在补偿精度和速度上都优于其他方法。同时利用MATLAB的图形用户界面环境(GUI)设计了湿度传感器的SVM温度补偿软件,并在该平台上进行了实例仿真,充分验证了软件的有效性。

Considering the humidity sensor easily affected by temperature can lead to the decline of the measurement accura- cy, improved AMPSO-SVM for temperature compensation of sensor was proposed, and compensation model was established, which was compared with the traditional GRID-SVM and GA-SVM. The results show that the improved algorithm can effectively reduce the effect of temperature, improve the measuring accuracy of humidity sensor, and is superior to other methods on the compensation accuracy and speed. At the same time, graphical user interface of MATLAB is used to design the temperature compensation software of humidity sensor. An example was simulated on the platform, thus validating the effectiveness of the software.