某型装甲车辆排烟红外辐射试验研究

Experimental Study on Infrared Radiation of Exhaust Gas From an Armored Vehicle

建立了线性标定和非线性标定数学模型,对不同温度的校准源在不同标定方法及不同标定策略测试的结果进行了分析。利用相对误差较小的标定方法和策略进行了试验,利用阈值函数筛选出了有效样本数据,经过充分训练和验证,建立了基于径向基神经网络的光谱辐射亮度校正模型,最后搭建了装甲车辆排烟红外辐射测试系统,得到了排烟的光谱辐射亮度。结果表明,针对不同的目标温度,采用相应的标定方法和策略能够减少红外辐射测试仪标定带来的测量误差;装甲车辆排烟利用测试数据建立径向基神经网络光谱辐射亮度校正模型能够有效降低大气吸收对测试结果的影响,获得更加准确的排烟红外光谱辐射亮度,该方法具有工程应用价值。

In order to solve the problem of large error in the test results of the infrared radiation characteristics of the exhaust gas of armored vehicles,the linear calibration and nonlinear calibration mathematical models were established by analyzing the infrared radiation measurement principle and the calibration principle of the infrared spectrum tester.The test results of the calibration sources with different temperatures were analyzed when using different calibration methods and different calibration strategies.The calibration method and strategy with small relative error were used to test,and the effective sample data were screened by threshold function.After full training and verification,the spectral radiance correction model based on Radial Basis Function(RBF)neural network was established.Finally,the infrared radiance test system of armored vehicle smoke exhaust was built,and the change of spectral radiance of smoke exhaust was obtained.The results show that,according to different target temperature,the calibration method and strategy can reduce the measurement error caused by the calibration of infrared radiation tester;the RBF neural network spectral radiance correction model based on the test data of exhaust gas of armored vehicles can effectively reduce the influence of big gas absorption on the test results,and then obtain more accurate infrared spectral radiance of exhaust gas.This method has engineering value.