Innovative Intelligent Methodology for the Classification of Soil Salinization Degree Using a Fractional-Order Master-Slave Chaotic System

Abstract

Soil salinization has become a highly significant eco-system issue that is encountered all over the world. Serious soil salinization leads to soil deterioration and has a negative impact on sustainable development of the eco-system and agriculture. However, the spectral reflectance of soils with high overlap and indecipherability makes it difficult to classify the soil salinization degree quickly and accurately. In this paper, an innovative, intelligent methodology using a fractional-order chaotic system to classify the soil salinization degree is proposed. To select a suitable order for the fractional-order chaotic system, the integer-order and noninteger order master-slave Lorenz chaotic systems were used to observe variations in the phase plane distributions. Movement traces of the chaotic system show that severely saline soil will exhibit more active changes, and its distribution status of the Lorenz chaotic system will be more scattered. After analyzing the characteristics of phase plane distributions, a preferred 0.9 fractional-order chaotic system is selected to obtain good analytical characteristics. Finally, extenics theory is used to verify the accuracy of salinization status classified by the coordinate values of the chaotic attractors, and an extenic matter element model is established to analyze the salinization degree. From the results, it was found that 100% analysis accuracy in the judgment of salinization level could be achieved under noninteger order status, and this judgment method is also suitable for soils in different human activity areas. This method has now become a benchmark for testing soil salinization with a chaotic system and is an innovative method that can be used to test the soil salinization degree quickly and accurately.