Paper Title
Prediction Of Cooling Load For Sericulture Using Artificial Neural Networks

- India continues to be one of the leading country in silk production. India stands second largest producer of silk. Sericulture has been developed to a precise science over years. In sericulture, thermal conditions plays a major role in the yield of cocoon cultivation. During incubation period of sericulture, the temperature should be maintained at 25O C with Relative Humidity (RH) 75%-80%, to maintain this manual setting has been done with the help of the air conditioner. During other moulting periods the temperature is maintained from 28O C to 24O C with RH value 85% to 70%. The change in the temperature along with RH has pronounced effect on moulting period. The temperature for spinning and cocoon preservation is maintained at 25O C. The resistance to high temperature is a heritable character and it may be possible to breed silkworm races tolerant to high temperature. Hence it is found that low temperature is always better than high temperature with reference to productivity of silkworm and larval duration for the different stages. This temperature maintenance is done manually by the workers. Thus we propose a predictive model, designed for automizing the control of temperature and humidity requirements of silk worm during various stages. According to the cooling load calculation, refrigeration system is designed. This paper is the analysis of regression which is the methodology adopted in Neural Networks. In this paper, a neural network (NN) model was developed to predict the cooling load of different stages of sericulture and to understand the causes of yield variability. First, we developed a NN model by relating stage type, time, outside temperature of shed, inside temperature of the shed and input humidity and evaluated model predictions for cooling load. We also explored the potential use of NN for predicting the cooling load. Finally, we evaluated the ability of the NN to attribute yield losses due to variations in thermal condition. Keywords - Artificial neural network, Precision, Sericulture, relative humidity, conditioner, cooling load.