遗传算法在智能气体检测装置中的应用研究

Study of genetic algorithms on intelligent gas test device

当今国内外对智能气体检测装置的研究异常活跃，它涉及传感器融合技术、计算机技术和应用数学以及各具体领域的一门综合性技术。其中最重要最难的因素是传感器特征提取技术和计算机特征数据融合技术，使智能气体检测装置能快速准确的识别不同气味。然而，目前尚无令人满意的方法。本文中研制了适用于传感器阵列反应的试验装置，在获得传感器与溶液气体反应的整个过程的数据的基础上，提取了传感器与溶液气体反应的特征值，利用遗传算法提出了一种新方法称为“遗传特征参数法”它能较好的识别不同溶液的气味。该方法也可用于解决其它模式识别问题。

The intelligent gas testing device has been well studied nowadays in home and aboard . It is a multi-technology that include the combination of sensor, computer, application maths and the field it used. The most important and difficult was how to get useful parameters from the sensor and how to use computer to combine the parameters. Which can make the intelligent device can identify the gas rapidly and accuracy. However, there was not a comfortable method. For this reason we developed a new testing device which satisfied the gas sensor array act with the gas,. The parameters of the sensor act with gas were picked up after getting the whole acting process data. We studied a new genetic algorithms which called genetic algorithm parameters . The new method can also be applied to other pattern recognition problems.