Earthquake Prediction Model Based on Neural Network and Gray Prediction Theory
钰淇 单 · Advances in Applied Mathematics · 2018
地震是一种威力巨大的自然灾害，强震的发生给社会带来重大灾难，威胁人类和平稳定的生活，令人类的文明发展遭受威胁，因此，对地震进行准确预测就十分必要。本文通过神经网络和灰度预测方法，分别建立地震的长期和短期预测模型，通过模型对地震的发生情况进行科学预测，以减少地震带来的损失。首先，根据强震带分布，对近50年来世界地震数据进行初步筛选，就地震活动的随机性较强这一特点，本文选用神经网络建立地震长期预测模型。通过模型分析中国历年地震活动次数及其震级强度，验证了模型预测准确、效果好的特点。进一步，我们应用此模型对中国2018年地震发生情况进行预测。之后，使用灰度预测理论建立短期地震预测模型，研究十年内发生地震的情况，分析发生地震的时间以及震级强度两组数据，建立地震短期灰度预测模型。通过对比以往数据，我们发现灰度预测模型能够很好地拟合震级大小，且误差范围小，预测效果理想。 Earthquake is a powerful natural disaster. The occurrence of earthquake has brought great disaster to human society. It threatens both the peaceful life and the development of civilization of human being. Therefore, it is necessary to predict the earthquake accurately. In this paper, a long-term prediction model of earthquake was established based on neural network and a short-term prediction model of earthquake was established based on gray prediction theory. The models are used to predict the occurrence of earthquakes in order to reduce the earthquake disaster loss. Firstly, according to the distribution of global pleistoseismic zone, the last 50 years’ seismic data has been preliminarily screened. In this paper, neural network was used to establish long-term earthquake prediction model due to the randomness of seismic activity. It was proved by analyzing the number of times and magnitude of earthquake activity in China that the predicted model was exact. Further, this model was applied to predict the occurrence of earthquake in China in 2018. Then, we analyzed the time and the magnitude of earthquake in the last ten years. Based on gray prediction theory, we established short-term prediction model of earthquake. The accuracy of model was verified by analyzing former data. The gray prediction model can fit the magnitude of the earthquake well.