Hybrid neural network classification for irrigation control in WSN based precision agriculture

Decision support systems (DSS) were built using the support of wireless sensors network (WSN) for resolving many real-world issues. Precision agriculture (PA) is the most popular area which requires DSS. Numerous agricultural cropping schemes in arid and semiarid areas practice irrigation process which is a crucial one and also here the main concern is water applications and management. An automatic Smart data mining based Irrigation Support Scheme is projected in our work in order to manage the irrigation in agriculture. Then for irrigation management, the author introduced the work Convolutional Neural Support Vector Machines Hybrid Classifier (CNSVMHC). This, in turn, avoids the weekly irrigations which is required for plantation. In this proposed research work, real time soil moisture content (MC) data collection were performed with the assistance of WSN and then irrigation will be controlled according to those collected data through CNSVMHC for an efficient irrigation management. The CNSVMHC is a heterogeneous combination of the convolutional neural network (CNN) and support vector machines (SVM), where the output layer of the CNN is substituted by an SVM. A control system with closed loop scheme was enabled through this process, which adjust the decision support scheme to approximation faults and local perturbations. As of the intricate and varied information dependent systems, the effectiveness and consistency of irrigation can be preserved through the soil, weather, and water and crop data. In order to do this process, we need help from the sensor network and other agricultural techniques for storing and using the rain water, maximizing their crop productivity, minimize the cost for cultivation and utilize the real time values rather than depending on prediction.