ABSTRACT Drought is one of the most recurrent natural disasters with cataclysmic effects on water budget, crop production, economic progression and public health. These consequences are magnified by the climate change leading to more intense drought conditions. A number of drought indices have been presented to calibrate the drought severity with its own strengths and limitations. Many of them are region-specific and unable to exhibit the alterations in significant drought inducing elements. Internet of Things (IoT) is well-suited for continuous monitoring, collection and analysis of different environmental phenomena. The dimensionality of the data collected about drought inducing attributes temperature, humidity, precipitation, evapotranspiration, groundwater, soil moisture at different depths, streamflow and season is reduced using PCA (Principal Component Analysis) at fog layer. Cloud layer estimates the drought severity level using Artificial Neural Network (ANN) whose parameters are optimised with Genetic Algorithm (GA) to get more accurate system and ARIMA method is used to forecast the drought for different time frames. Experimentation done on data collected from government websites shows that proposed system performs well in terms of accuracy, sensitivity, specificity, precision and F-measure with values 95.03%, 90.6%, 96.73%, 91.42% and 91.01%.

中文翻译：

---

使用人工神经网络和遗传算法的基于云雾的干旱预测和预报框架

摘要 干旱是最常发生的自然灾害之一，对水资源收支、作物生产、经济发展和公共卫生产生灾难性影响。导致更严重干旱条件的气候变化放大了这些后果。已经提出了许多干旱指数，以根据其自身的优势和局限性来校准干旱的严重程度。其中许多是区域特异性的，无法表现出显着干旱诱导因素的改变。物联网 (IoT) 非常适合对不同环境现象进行持续监测、收集和分析。收集到的干旱诱导属性数据的维度包括不同深度的温度、湿度、降水、蒸散、地下水、土壤水分、在雾层使用 PCA（主成分分析）减少流量和季节。云层使用人工神经网络（ANN）估计干旱严重程度，人工神经网络（ANN）使用遗传算法（GA）优化参数以获得更准确的系统，ARIMA方法用于预测不同时间范围的干旱。对从政府网站收集的数据进行的实验表明，所提出的系统在准确度、灵敏度、特异性、精密度和 F 值方面表现良好，分别为 95.03%、90.6%、96.73%、91.42% 和 91.01%。