灌区土壤盐渍化程度云理论改进多级模糊评价模型

土壤盐渍化的产生过程是一个多指标参与、多层次驱动的复杂系统,针对盐渍化程度的评价的不确定性和模糊性问题,将不确定性云理论引入到土壤盐渍化程度的多级模糊评价中,通过构建土壤盐渍化程度多级模糊评价指标体系,构建了盐渍化程度的评语集云模型、诱发因子的隶属度云模型及权重云模型,进而提出了基于云理论改进的土壤盐渍化程度的多级模糊评价模型。同时,选定景泰川电力提灌灌区为研究区,对该灌区的土壤盐渍化程度进行了评价,并将评价结果和评语集云模型结合用Matlab仿真显示。研究表明:该灌区土壤的盐渍化处于轻度盐化土和中度盐化土之间,0~100 cm土壤的含盐量为0.224 2%的可能性最大;利用云理论改进多级模糊评价模型对土壤的盐渍化程度开展研究,用不确定性云参数代替精确数值,更具普遍性。相关研究可为开展盐渍化程度评估和预测的研究提供有益参考。

Improved multi-level fuzzy evaluation model based on cloud theory for evaluation of soil salinization degree

Abstract: The producing process of soil salinization is a complex fuzzy system with participation of multi indices and multi-level driving, which involves field water transformation, field heat exchange and soil salt-water transport. Therefore, in view of this fuzzy process, the comprehensive evaluation model of salinization based on multi-level fuzzy theory can be used to reveal it quantitatively. However, this model can take good account of the fuzziness of the complex system in the comprehensive evaluation of soil salinization degree, the randomness and discreteness of the system are neglected yet, and the fuzzy nature of the randomness and volatility of fuzzy systems are not well represented. The cloud theory can describe the randomness and uncertainty of fuzzy systems well, in which qualitative concepts and quantitative values can be freely transformed, and the subjective and individual empirical effects of experts in describing the status of inducing factors of soil salinization can be well avoided. In view of this, in this paper, the uncertainty cloud theory was introduced into the multi-level fuzzy evaluation model, and the driving process of soil salinization was divided into 4 layers: The evaluation layer, driving process layer, inducing factor layer and element status layer. Regional soil salinization degree was described, the multi-level fuzzy evaluation index system of soil salinization degree was constructed by the basic principles of analytic hierarchy process and multilevel fuzzy theory, and a set of cloud model of salinity evaluation was constructed by using normal cloud generator. Meanwhile, the scale criterion of inducing factor of soil salinization based on cloud scale was constructed by improving the traditional Satty scaling principle, and a weight cloud model of induced factors was constructed. In addition, the membership cloud model of the induced factor was constructed by using the backward cloud generator. Finally, the weight cloud model and membership cloud model were weighted to determine the evaluation model of soil salinization degree, and then a multi-level fuzzy evaluation model of soil salinization based on cloud theory was proposed. Moreover, the model was used to evaluate the degree of soil salinization of Jingtaichuan electric pumping irrigation area, Gansu Province. And then the evaluation results and comments collection cloud model were combined, which was for emulation display by MATLAB software. The result shows that: 1) The salinization degree of soil in irrigated area is between slight and moderate. The expected value of soil salinity is 0.224 2%, that is to say, the likelihood of 0.224 2% soil salinity in 0-100 cm is maximum. In addition, the entropy and hyper entropy of the cloud model are 0.029 5 and 0.021 2, respectively, and the value is smaller, that is, the uncertainty of the evaluation results is small, and the evaluation results fluctuate in a small range. The evaluation results basically conform to the actual situation of irrigation area, and the evaluation results are good, which verify the feasibility of the model. 2) The multi-level fuzzy evaluation model is improved by using cloud theory, the stability and reliability of the results are also given besides the expected values, and the fuzziness, randomness and discreteness are organically combined by 3 numerical characteristics of the cloud model i.e. expectation value, entropy and hyper entropy. Compared with the multi-level fuzzy evaluation model, the results are more in line with human language habits, and the information is more abundant, which provides a new method for the evaluation of soil salinization degree in irrigation area.