CAJ | 학술논문

下水道可燃气体分析是城市下水道可燃气体监测预警系统的重要组成部分。该文针对BP神经网络对下水道可燃混合气体分析存在速度慢、容易陷入局部最优，以及标准最小二乘支持向量机鲁棒性差的缺点，建立了一种基于加权最小二乘支持向量机（WLS-SVM ）的下水道可燃气体分析模型。加权最小二乘支持向量机模型采用最小二乘线性系统，对误差变量进行权值设定，提高了学习速度和学习精度。仿真结果表明：基于WLS-SVM的下水道可燃气体分析模型优于所比较的BP神经网络和最小二乘支持向量回归机2种模型，具有优良的预测精度和鲁棒性。
하수도가연기체분석시성시하수도가연기체감측예경계통적중요조성부분。해문침대BP신경망락대하수도가연혼합기체분석존재속도만、용역함입국부최우，이급표준최소이승지지향량궤로봉성차적결점，건립료일충기우가권최소이승지지향량궤（WLS-SVM ）적하수도가연기체분석모형。가권최소이승지지향량궤모형채용최소이승선성계통，대오차변량진행권치설정，제고료학습속도화학습정도。방진결과표명：기우WLS-SVM적하수도가연기체분석모형우우소비교적BP신경망락화최소이승지지향량회귀궤2충모형，구유우량적예측정도화로봉성。
The analysis of Sewer combustible gas is an important part of city sewer combustible gas monito-ring and warning system .For the disadvantages of slow convergence and easy to fall into local optimum by means of BP neural network to analysis of sewer combustible gas mixture ,a sewer combustible gas analy-sis model based on weighted least squares support vector machine (WLS-SVM ) has been established .By means of RBF kernel function ,the weighted least squares support vector machine model effectively reduced the complexity and improved the learning speed .The simulations reveal that the sewer combustible gas a-nalysis model based on the WLS-SVM is superior to the other two models ,BP neural network and LS-SVM ,with good generalization performance and prediction accuracy .