CUMT-UCASAL Joint Research Center for Biomining and Soil Ecological Restoration, State Key Laboratory of Coal Resources and Safe Mining, China University of Mining and Technology, Xuzhou, Jiangsu, China.
State Key Laboratory of Coal Resources and Safe Mining, China University of Mining and Technology (Beijing), Beijing, China.
PLoS One. 2021 Oct 28;16(10):e0259155. doi: 10.1371/journal.pone.0259155. eCollection 2021.
The development of "CC30A CH4-CO2 combined analyzer" with infrared gas sensor as the core detection device can be widely used in online gas component analysis. In data analysis, the maximum value and arithmetic mean of the sensor data for each test period are not effective value. The characteristics of the dynamic data are: (1) Each DAW completes one test for one parameter, there is a unique effective value; (2) In test state, the fluctuation of the sensor value gradually decreases when approaching to the end of the test. An effective value calculation model was designed using the method of dimensionality reduction of dynamic data. The model was based on the distribution characteristics of the process data, and consists of 4 key steps: (1) Identify the Data Analysis Window (DAW) and build DAW dataset; (2) Calculate the value of optimal DAW dataset segmentation and build DAW subdataset; (3) Calculate the arithmetic mean (Mc) and count the amount of data in each subdataset (Fc), and build the optimal segmentation statistical set; (4) Effective value calculation and error evaluation. Calculation result with 50 sets of monitor data conformed that the EVC model for dynamic data of gas online monitoring meets the requirements of experimental accuracy requirements and test error. This method can be independently calculated without relying on the feedback information of the monitoring device, and it has positive significance for using the algorithm to reduce the hardware design complexity.
以红外气体传感器为核心检测器件的“CC30A CH4-CO2 组合分析仪”的研制，可以广泛应用于在线气体成分分析。在数据分析中，每个测试周期内传感器数据的最大值和算术平均值都不是有效值。动态数据的特点为：（1）每个 DAW 为一个参数完成一次测试，有唯一的有效值；（2）在测试状态下，当接近测试结束时，传感器值的波动逐渐减小。利用动态数据降维的方法，设计了一种有效值计算模型。该模型基于过程数据的分布特征，由 4 个关键步骤组成：（1）识别数据分析窗口（DAW）并构建 DAW 数据集；（2）计算最佳 DAW 数据集分段的值并构建 DAW 子数据集；（3）计算算术平均值（Mc）和每个子数据集的数据量（Fc），并构建最佳分段统计集；（4）有效值计算和误差评估。用 50 组监测数据进行计算结果表明，气体在线监测动态数据的 EVC 模型满足实验精度要求和测试误差要求。该方法可以独立计算，无需依赖监测设备的反馈信息，对于利用算法降低硬件设计复杂度具有积极意义。
