Environmental Control Decision Method for Livestock and Poultry House Based on Neural Networks and Improved D-S Evidence Theory

摘要

In order to improve the accuracy of environmental control decision in livestock and poultry breeding, an environmental control decision method was designed, based on neural network and improved Dempster-Shafer (D-S) evidence theory. Firstly, the outlier data, collected by sensor, was detected and repaired using box chart method and mean substitution method. Secondly, the first-level decision classification of the data was carried out by using the neural network SoftMax multi-classification method. Finally, the improved D-S evidence theory algorithm was used to implement the final decision of the environmental control system of livestock and poultry house. Taking the environmental control system of chicken house as an example, the simulation experiment verifies the proposed method. The experimental results show that the method can make decision accurately on the environmental control system of chicken house. The proposed method has certain promotion and application value.