人工神经网络在生烃条件评价中的应用

根据生烃条件评价的地质模型,用人工神经网络建立权重分配,改变了以前人为给定权值的方法。通过网络的自学习,自动地建立起权重分配关系,使权重分配较为客观准确。用该方法对鄂尔多斯盆地奥陶系生烃条件进行评价,取得了与专家系统评价基本一致的结果。

APPLICATION OF AN ARTIFICIAL NEURAL NETWORK TO EVALUATING CONDITIONS OF HYDROCARBON GENERATION

According to the geological model for evaluating condition of hydrocarbon generation,an artifical neural network with three-layer networks,multi-imput and single output isbuilt. The hidden layer consists of four neural cells'This neural network can acquire weightby itself, and modify constantly the connections between the cells. The weight-assigned connections are automatically built by the artifical neural network, thus renewing the old methodof manned weight-assignment. This method is more objective and accurate.The method isused to evaluate the hydrocarbon generation condition of the Ordovician System in OrdosBasin, the result is basically the same as that of the expert system.