基于离群模糊核聚类算法的PID毒气检测系统设计

Design of PID Toxic Gas Detection System Based on Outlier Fuzzy Kernel Clustering

摘要:

针对传统毒气检测系统混合检测中适用性差、检测误差率高的不足，提出基于离群模糊核聚类算法的PID毒气检测系统设计。在系统硬件设计中选择了性能更强的STM32F2X型MCU，并设计了专门用于毒气类别分析的功能模块；在软件算法和主控程序的设计中，采用了离群模糊核聚类算法提高对毒气数据的聚类分析能力，以此改善毒气检测的准确性。实验结果表明，提出的PID毒气检测系统能够识别出多种天然毒气和化学毒气，在毒气浓度的检测误差方面也能够控制在2%以内。

Abstract:

To overcome the disadvantages of poor applicability and high detection error rate of traditional toxic gas detection system, the design of PID toxic gas detection system based on outlier fuzzy kernel clustering algorithm is proposed. In the hardware design of the system, STM32F2X MCU with stronger performance is selected, and the function modules specially used for toxic gas category analysis are designed. In the design of software algorithm and main control program, outlier fuzzy kernel clustering algorithm is used to improve the clustering analysis ability of toxic gas data, so as to improve the accuracy of toxic gas detection. The experimental results show that the proposed PID toxic gas detection system can identify a variety of natural and chemical toxic gases, and the detection error of toxic gas concentration can be controlled within 2%.