ANN based modeling and control of GHS for winter climate

Abstract

Most of the spices like hot chilli, pepper are usually grown as annuals. Hence, they need to survive in both summer climate as well as winter climate. Commonly, the productivity of hot chilli plant will be affected by cold climate conditions. To protect the hot chilli plant from winter temperature greenhouse heating system is required. Modelling and control of nonlinear Green House System (GHS) are cumbersome due to strong interactions between physical phenomena and biological systems. The heating and ventilation are the important factors in designing greenhouse for the plants to grow in winter climate. The GHS is regulated with respect to temperature and humidity using heating and ventilation systems. _In this paper, Artificial Neural Network (ANN) based Nonlinear Auto Regressive with Exogenous input (NARX) with time series model is developed for the winter climate of GHS with the temperature about 32 °C to 35 °C and humidity 12 g/kg to 8g/kg. Using this model, a Nonlinear Auto Regressive Moving Average (NARMA-L2) controller is designed to obtain the desired closed loop performance to achieve enhanced productivity and quality of hot chillies.