Performance prediction of solid desiccant rotary system using artificial neural network

1st International Conference on Contemporary Research in Mechanical Engineering with Focus on Materials and Manufacturing

机械制造;材料科学

Mishra, V.K.^1 ; Singh, R.P.^1 ; Das, R.K.^1

Indian Institute of Technology (ISM) Dhanbad, India^1

关键词: Air inlet temperature;    Artificial neural network modeling;    Correlation factors;    Desiccant wheels;    Experimental test;    Performance prediction;    Regeneration temperature;    Temperature and relative humidity;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/404/1/012006/pdf
DOI  :  10.1088/1757-899X/404/1/012006

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

This paper presents an artificial neural network model for solid desiccant rotary system to predict its performance in terms of temperature and relative humidity of process air leaving the desiccant wheel after losing latent heat. Present paper also explains the experimental test setup that is used for taking reading. The experimental readings are all taken at steady state by varying the input conditions such as process air inlet velocity, regeneration air inlet velocity and regeneration temperature and process air inlet temperature and relative humidity. Majority of data taken from experiments is used to train the model (85%) and rest (15%) is used for testing of the model. The performance output predicted by the ANN model have high correlation factor(R>0.98336). The results predicted by the ANN model shows that ANN model can be successfully applied to predict the performance of solid desiccant wheel with sufficient accuracy and reliability.