Development and experiment of intelligent monitoring system for geese house environment
guo binbin,sun aidong,ding weimin,shi zhendan,zhao sanqin,yang hongbing
DOI: https://doi.org/10.11975/j.issn.1002-6819.2017.09.023
2017-01-01
Abstract:Out-of-season breeding technology for goose was proposed due to their obviously seasonality of production.Precisely time control is the most critical factor in this technology.Simultaneously,temperature and humidity control are also needed in case of heat stress.The application of the new technology brings challenges to the traditional breeding methods.Environment control in goose house was mostly artificial or semi mechanized.Manual operation was of poor real-time performance,and could not achieve the organic connection between the various environmental factors.Light,temperature,relative humidity and other environmental parameters could not be recorded real-time.When diseases or some abnormal reproduction happen,managers could not trace causes from the aspect of environment.This study therefore was conducted to develop an intelligent geese house environment monitoring system for assisting geese out-of-season breeding practices.In the system,BP neural network was utilized to construct an intelligent control model with history temperature and relative humidity data.With this model,the system could make comparison between current temperature and relative humidity inside and outside,so as to determine the number of working fans and the on-off state of cooling pad.GPRS wireless transmission module was used to transmit in-house environmental parameters.With this system,mobile terminal GSM function was also utilized to remote control the geese house equipment such as fans,illuminating lights,pumps and cooling pads.On the basic framework technology of EXT,Hibernate and Spring,a J2 EE enterprise web application was built.This made it possible of remote control and real time monitoring.In addition,users could also obtain the environment conditions from mobile client or remote control the corresponding equipment.Field tests were conducted in comparison with artificial control and other commercially developed animal house monitoring systems.The system had been running for more than five months.Results showed that this system was stable and reliable during long-term operation,and could meet the requirements of light,temperature and relative humidity of out-of-season breeding.Compared with artificial control and Shanghai Fanlong controller,the control precision was improved by 5.49%and 2.83%,respectively,labor cost was cut down by 50%.Geese elimination rate was decreased by 1.09%and 0.62%,respectively.With the method of cooling pads system in summer,the RMSE between in house temperature of BP neural network(A3),artificial experience(A2),setting static operating point(A4) and the set values were 0.202,0.494 and 0.372℃,respectively,and relative humidity was 1.745%,3.166%and 2.621%,respectively.Under the precise control of light,geese in all three houses exhibited normal onset and level of egg laying.They began to lay eggs from May 1st,and rose to the peak level in about one month.During hot summer,there was no heat stress of geese in A3,they showed steady laying rate,while laying rate of geese in A2 and A4 showed varying degrees of decline.At the peak of egg production,all three groups maintained their laying rate between 35%and 45%.From May 1st to October 1st,the mean daily egg laying rate of A2,A3 and A4 are 24.77%,31.17%and 27.94%,respectively,which indicated that the intelligent control system of geese house environments allowed geese to exhibit normal out-of-season production performances.