Smart Greenhouse Monitoring System Using Internet of Things and Artificial Intelligence
Sultana Jahan Soheli,Nusrat Jahan,Md. Bipul Hossain,Apurba Adhikary,Ashikur Rahman Khan,M. Wahiduzzaman
DOI: https://doi.org/10.1007/s11277-022-09528-x
IF: 2.017
2022-01-21
Wireless Personal Communications
Abstract:Climate change has already proven its terrible effect on agriculture. Although greenhouse is already an established system for crop production, with technological advancement it is possible to apply automation in many parts of this greenhouse. Therefore, an automated smart greenhouse based on an adaptive neuro fuzzy inference system (ANFIS) and Internet of Things (IoT) could be the best solution to boost the crop production inside the house. Where, four kind of weather data such as temperature, humidity, sunlight and soil-moisture are being collected by using sensors in real time. These collected data are then feed as input variables to the fuzzy control system. The fuzzy control system manipulate the data and ANFIS then make prediction for optimum values of the weather parameters. Thus farmers can monitor all the data and can decide the best value for temperature and humidity. The end users (farmers) can visualize all the data by a simple mobile app installed on their cell phone. GSM or TCP/IP is being used for all kind of data transferring. The FIS node also utilizes same networks to transfer IoT perception layer data to application layer. To ensure the data security, four types of potential IoT perceptron layer attacks are considered and shown their probability to occur through the confusion matrix. Later, necessary steps are taken to prohibit the attacks. Here winter crops are considered in the final simulation, when the optimum temperature in winter is 24º Celsius and humidity is 76.00%. The system is 93.62% capable to detect any attack or security breach at perception layer with a Precision value of 0.83, recall of 0.78 and FI score is 0.81. In comparison to other recently proposed and available systems, this work also combines IoT technology for identifying data threat on a network transfer with fuzzy set. This approach improves learning efficiency, improves prediction accuracy, and proved to be a feasible and effective automated greenhouse maintenance system. Simultaneously, the data collecting module and presentation schema of data from various sensors, as well as the security subsystem module, achieve cloud data storage and format conversion that is compliant with protocol format data. As a result, it may provide data traceability and durability for customized indoor agriculture quality and safety. Thus this modern greenhouse maintenance system is efficient, cost effective, secure and easy to use.
telecommunications