Weigh-in-motion Of The Dairy Cow Based On STM32 Microcontroller

The weight of dairy cows is an important indicator of its growth and development,which is an important reference data for the precision cows breeding.It is the basis for measuring the production performance,determining the feed intake,the nutrient supply,the first mating age,the dosage,the seed selection and so on.Therefore,measuring the weight of dairy cows is the regular work for cattle feeding.But at this stage animals are weighted through the weighbridge with fence or their weight data are assessed by body size.These methods consume worker’s more energy and weighing cows too frequently will stimulate dairy cows,cause yield reduction or even no yield.Some professional large-scale dairy farms introduce Israel’s AfiMilk Pasture Management System or DeLaval Automatic Weighing System for Cows to solve that problem.These systems are complex and difficult to maintain,the basic after-sales could not be guaranteed.If breeders and managers’ cultural level can’t reach required standard,the weighing data of dairy cows will be typically underutilized and the real-time monitoring system of weighing cows can’t be implemented.Most of the automatic weighing equipment introduced from abroad have gradually fallen into disuse.The purpose and the meaning of this study is weighting dairy cows once a day without any stimulation to them and can easily cooperate with other pasture management software.The main content of this paper is as follows:1.The specific requirements of the system is determined by scene researching and searching the literature,According to the laws of the cows walk characteristics and daily activities.Then the overall design of the system is developed.2.The hardware implementation of Weigh-in-Motion System mainly includes the construction of the weighing platform and the design of the weighing system circuit.The weighing platform includes mechanical components,pressure sensors,infrared switches,eartag recognizer and so on.The weighing system circuit includes a signal processing module,a MCU module and a wireless communication module.3.The system software program design is divided into MCU software program design and upper computer management software program design.MCU software program is mainly divided into seven parts as following: main program part,infrared switches control program,SCADA system program,data-processing system program,display program and the program stability.Upper computer management software is mainly weight data management software for the dairy cows.4.Reliability verification of the system.After the completion of data acquisition and data preprocessing,walking characteristics and the sampled data-regulation are analyzed.Cows usually have three walking states,walking with pause,smooth walking or walking with stimulation.Three mathematical processing algorithms have been studied separately for the three states.The static weighing experiment and the dynamic weighing experiment of dairy cows were completed and the error of weighing in various walking states has been got.The feasibility and reliability of this system are verified through experiments.To further improve the weighting accuracy of this system,and reduce its weighting error,BP neural network is proposed in this study.The data parameters of the four weighing sensors are considered as inputs to BP neural network to finish data prediction for dairy cows body weight.BP neural network train through a large number of experimental weighing data could reduce the weighting error.After initial lots of tests,this system could achieve the purpose of dynamic weighing,and the automatic weighing system is done during daily activities without any stimulation of cows or even without any extra labour.This system makes more efficient using of weighing data cooperate with farm management software through host computer.It is one of the most important part in automatic and precision farming and it has more broad prospects.