Cowpea Hot Air Drying Moisture Prediction Model Based On Elman Neural Networks

Cowpea contains rich nutritional value, but its high respiration rate, so it is hard for a long time storage of fresh products. Cowpea is made fresh dried products, not only to save most of the nutrients, to extend the storage time, but also to increase the taste. Drying and processing of fresh cowpea, has great significance for cowpea production and sales.Study This paper first hot-air drying characteristics of cowpea were discussed, drawing the drying curve. Cowpea and hot air drying process is analyzed and discussed.The study found drying in a constant temperature, air temperature, hot air drying performance speed on the larger, the smaller the impact cowpea length; in variable temperature drying, the initial temperature, the end of the hot air temperature and speed are the three factors that affect the performance of cowpea dry.Then by varying the temperature dry orthogonal experiment to rehydration ratio and drying rate for test indicators, identify the primary and secondary factors of the order.Factors that impact on the ratio of the late rehydration temperature> initial temperature> hot air speed; factors affecting the drying rate of the late temperature> hot air speed> initial temperature, to obtain the optimum process parameters for the initial temperature 40 ℃, the end of the temperature 65 ℃, hot wind speed 1.8m/s.At last using Elman neural network to build cowpea variable temperature drying process moisture prediction models, and the model for network training and testing. The model training sample mean square error is 0.0010153, the average relative error of the test sample was 3.41%. The results show that the model predicts high precision, high fitting degree, variable temperature drying process for cowpea moisture content prediction is feasible and effective.