摘要
为了实现对煤与瓦斯突出的快速、准确和动态预测,考虑煤与瓦斯突出的多种影响因素,提出了基于AdaBoost增强学习和逻辑回归(LR)的煤与瓦斯突出预测方法。在AdaBoost的框架下,进行LR多次学习,完成煤与瓦斯突出预测模型构建;加入学习率和正则化参数进行控制,弱化学习效果,防止过拟合;基于反馈的实际结果,完成煤与瓦斯突出预测模型的修正。根据实例验证结果,构建的煤与瓦斯突出预测模型建模过程稳定、准确性高、建模消耗时间短、满足实时要求,证明了提出的煤与瓦斯突出预测模型构建算法是有效的。
In order to realize the accurate,quick and dynamic prediction of coal and gas outburst,considering multiple influencing factors of coal and gas outburst,a prediction method based on AdaBoost reinforcement learning and logistic regression(LR)is proposed.In the framework of AdaBoost,the prediction model of coal and gas outburst is built after learning LR many times.The learning rate and regularization parameters are introduced to weaken the learning effects and avoid over-fitting.The prediction model of coal and gas outburst is updated based on actual feedback results.The results of example verification show that the prediction model of coal and gas outburst constructed in this paper has a stable modeling process,high accuracy,short modeling time,and meets the real-time requirements.Therefore,the proposed construction algorithm for prediction model of coal and gas outburst is effective.
作者
阎馨
吴书文
屠乃威
朱永浩
付华
YAN Xin;WU Shu-wen;TUNai-wei;ZHU Yong-hao;FU Hua(Faculty of Electrical and Control Engineering,Liaoning Technical University,Huludao 125105,China)
出处
《控制工程》
CSCD
北大核心
2021年第10期1983-1988,共6页
Control Engineering of China
基金
国家自然科学基金项目(61601212,71771111)
辽宁省教育厅辽宁省高等学校基本科研项目(LJ2017QL012)
辽宁省教育厅重点实验室项目(LJZS003)
辽宁工程技术大学博士启动基金项目(14-1102)。
