| In recent years,with the continuous development of economy and the continuous improvement of the Engel coefficient,people’s requirements for quality of life have gradually increased.According to the data,it shows that the country is attending to integrate modern science and technology into livestock breeding,due to the rapidly growing demand for meat products by Chinese residents,and the country’s strong advocacy of the development of green,healthy and sustainable breeding in recent years.At present,China’s breeding industry is mainly based on small and medium-scale breeding,and the level of modernization is not much high.During the breeding process,farmers mainly follow their own working experience to regulate the breeding environment,and as to the monitoring and control of the breeding environment still need to be manually operated by going to the site.In the livestock farms,temperature,humidity,ammonia,carbon dioxide and other important environmental factors are related to the health of livestock,if these environmental factors are not monitored in a timely manner,it will directly lead to lower production efficiency.Therefore,the application of modern science and technology to traditional breeding industry can implement effective monitoring and advance regulation of the main environmental information of the farm,which directly affects the level of production efficiency of the farm and has important research value and broad application prospects.In response to the above problems,this thesis designs a monitoring system on pig farm environment based on ARM.This system can collect temperature,humidity,carbon dioxide concentration and ammonia concentration in the pig farm,and can control the environment in the pig farm according to the collected environmental data.At the same time,it can predict the environmental changes of the pig farm in the next time period by using long and short-term memory neural network(LSTM)to achieve the purpose of advance regulation.The development of the system is divided into three parts: hardware design,software design and environment prediction.In the hardware part,the acquisition node uses the STM32 to drive the sensor to collect the environmental data of the pig farm,uses the LoRa module to send and receive data,and uses the combination of the STM32,the LoRa module and the WiFi module to design the intelligent gateway module,which uses LoRa module to communicate with the acquisition terminal,and connects to the Internet and receives and sends information through WiFi module.In the software part,the multi-task scheduling mechanism of RT-Thread real-time operating system is used to ensure the orderly and stable operation of data collection,data transmission,alarm and other functions,and the OneNet IoT cloud platform is used to store and call the environmental data of pig farms,real-time monitoring of environmental information and remote control of hardware equipment in the pig farm.In the part of environment prediction,the long and short-term memory neural network is used to train the environment prediction model in the pig farm,and the prediction model is used to predict the environmental data of the pig farm in the next time period,so as to grasp the environmental changes in the farm in advance and make corresponding control measures in time.After the system was built,each functional module of the system was tested and verified.Through the test,the software and hardware of the system could work normally and the overall operation of the system could achieve the expected effect,because the prediction error of environment prediction model based on LSTM algorithm in the pig farm was within the expected range of the design.Ultimately,this system can collect various environmental data in pig farms accurately,and realize the advance regulation of the farm environment according to the collected data,and also predict the environmental data in the next time period by using LSTM algorithm,so that farmers can fully experience the intelligence and convenience of modern farming,which has the certain application value. |
