基于免疫机理的黑龙江省耕地系统安全预警研究

Study on Early-Warning of Cultivated Land System Security in Heilongjiang Province based on the Immune Mechanism

研究目的:科学识别黑龙江省耕地系统安全警情现状及其未来发展趋势。研究方法:改进突变级数模型和Elman神经网络模型。研究结果:(1)构建基于生物免疫机理的耕地系统安全警情评定指标体系,得出黑龙江省警情1995—2014年整体呈波动上升趋势,警度从"无警"上升至"轻警",再至"中警",说明基于改进突变级数模型的评价结果可凸显评价对象之"优、劣"特征;(2)引入Elman神经网络构建耕地系统安全警情预测模型,测试拟合精度在0.97以上,预测结果可靠,表明未来20年黑龙江省警情将整体呈先升高、后降低趋势,警度在"轻警"与"中警"之间。研究结论:基于生物免疫机理的黑龙江省耕地系统安全预警研究,可为省域尺度下耕地系统安全预警提供一种新的思路。

The purpose of this study is to scientifically identify the current situation and future trend of the cultivated land system security level in Heilongjiang Province. The research methods employed were subjective and objective weighting methods, advanced catastrophe progression model and Elman neural network. The results are showed as follows. 1） Based on the immune mechanisms of biological system, the cultivated land security level evaluation index system was constructed, finding that the alert level for Heilongjiang Province from 1995—2014 was generally on an upward trend. The alert level raised from ＆quot;nil＆quot; to ＆quot;light＆quot;, and then up to ＆quot;moderate＆quot; from 1995 to 2014, thus showing that the evaluation results based on the improved catastrophe progression model highlighted the ＆quot;excellent or inferior features＆quot; of the evaluation object. 2）Using the Elman neural network to construct the cultivated land system security early-warning model, the matching accuracy of measurements was greater than 0.97 and the prediction results were reliable. The security level in Heilongjiang Province will increase then decrease during the next 20 years, and the alert level will be between ＆quot;light＆quot; and ＆quot;moderate＆quot;. In conclusion, the research on cultivated land system security early-warning in Heilongjiang province based on the biological immunity mechanisms can provide a new way for cultivated land system security early-warning at the provincial level.