基于广义线性模型的概率风险评价方法及其应用

来源期刊:中南大学学报(自然科学版)2011年第6期

论文作者:张进春 吴超

文章页码:1719 - 1724

关键词:广义线性模型;概率风险评价;极大似然估计;皮尔逊χ2检验

Key words:generalized linear models(GLM); probabilistic risk assessment; maximum likelihood estimation(MLE); Pearson χ2 test

摘    要:应用广义线性模型处理风险评价中响应变量为属性变量或离散型变量的问题,对风险事件依风险等级的累计发生概率进行logistic变换,以变换的累计发生概率作为连接函数,基于风险样本事件先验信息建立回归模型。针对待评价的风险事件,利用所建立的回归模型分别计算各风险等级的累计发生概率和发生概率,以其最大发生概率所对应的等级作为风险事件的最终评价等级。在建模过程中,采用极大似然估计法对回归系数进行参数估计并采用Newton-Raphson迭代算法求解,模型的拟合优度采用皮尔逊χ2检验。应用本方法建立煤与瓦斯突出的概率风险评价模型,并对具体矿井煤与瓦斯突出的概率风险进行评价。研究结果表明:采用该方法可准确得出评价结果。

Abstract: The basic principle of generalized linear models(GLM) was used in risk assessment method. The cumulative occurrence probability of the risk was transformed into logistic form according to its level. The risk assessment regression model was founded based on the prior information of the samples taking the transformed cumulative probability as the linking function. Then every level’s cumulative occurrence probabilities of the risks assessed were gained using the founded regression model and the occurrence probability was calculated. The final risk level was determined with the level of maximum occurrence probability. The coefficients of the regression model were estimated with MLE and calculated by Newton-Raphson iterative algorithm, and the goodness of fit of the model was tested by Pearson χ2 test. The method was applied in coal and gas outburst risk evaluation and a specific mine was assessed. The results show that this method is accurate and valuable.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号