论文部分内容阅读
为了提高矿井上隅角瓦斯(这里指甲烷)浓度的预测精度,得到瓦斯浓度的一个预测范围,提出一种将模糊信息粒化(FIG)和支持向量机(SVM)相结合的瓦斯浓度预测方法。首先利用模糊信息粒化对原始数据进行模糊粒化处理,并且给出一个预测范围。然后将粒化后的数据作为输入,运用SVM进行回归预测,采用粒子群(PSO)算法选取最佳的核函数参数g和惩罚因子c。最后根据实测值与预测值的对比判断预测方法的可靠度。试验结果表明:每一个时间段瓦斯浓度的实测值基本都在预测范围内,说明该模型预测精度较高,有较强的实用性和较快的收敛速度。
In order to improve the prediction accuracy of gas (here methane) concentration in the upper corner of the mine and obtain a prediction range of gas concentration, a gas concentration prediction method combining fuzzy information granulation (FIG) and support vector machine (SVM) is proposed . Firstly, the fuzzy data granulation is used to fuzzy the original data, and a prediction range is given. Then, the granulated data is taken as input, SVM is used for regression prediction, and particle swarm optimization (PSO) algorithm is used to select the best kernel parameter g and penalty factor c. Finally, the reliability of the prediction method is judged according to the comparison between the measured value and the predicted value. The experimental results show that the measured values of gas concentration in each time period are basically within the predicted range, indicating that the model has higher prediction accuracy, stronger practicality and faster convergence rate.