论文部分内容阅读
为了更加准确的表达矿山岩层运动的复杂演化规律,用模拟生物繁衍的进化算法进行模型识别研究.采用树状层次结构设计方法,不需事先假定模型的具体函数形式,根据进化过程中的选择、交叉、变异等遗传算子操作,可得到自适应模型.在VC++6.0语言平台基础上,自主开发了基于进化算法的矿山岩层运动模型识别系统,该系统具有模型智能识别和预测功能.应用结果表明,进化模型对采空区顶板离层的非线性演化过程的规律表达比较准确,模型值与实测值的误差非常小,平均仅为3.99%;以树状结构表达的进化算法适合描述岩层运动的复杂特征,开发的软件系统运行稳定、可靠,模型识别精度较高.
In order to more accurately express the complex evolution law of the strata movement in the mine, the model identification research is carried out by using the evolutionary algorithm of simulated biological multiplication.Using the tree hierarchy design method, without assuming the concrete function form of the model in advance, according to the choice in the evolutionary process, Crossover, mutation and other genetic operators can be obtained adaptive model.On the basis of VC ++ 6.0 language platform, the mine rock movement model identification system based on evolutionary algorithm is independently developed, and the system has the function of model intelligent identification and prediction. The results show that the evolution model is quite accurate for the nonlinear evolution process of the gob roof separation, the error between the model and the measured value is very small, with an average of only 3.99%; the evolutionary algorithm expressed in tree structure is suitable for describing the rock formation The complex features of the movement, the software system developed is stable and reliable, and the model recognition accuracy is high.