Dynamic Weighing System Design And Research For Animals

Weight as an important part of animal life characteristics,weight data is widely needed in animal husbandry and aquaculture,wildlife protection,medical research,and so on.Traditional weighing methods will keep the animals stationary in the weighing platform firstly,and then weighing them several times.This static weighing method has several drawbacks,for example,it affect the animal’s normal activities in the weighing process,as damage to animals;and weighing always longer,it reducing the efficiency and enhancing the costs;the weighing results always by doing average,when there is accidental error,the error will be too large.As the result,Empirical Mode Decomposition(EMD)algorithm is applied to the dynamic weighing of animals,and the hardware weighing system solved the drawbacks of the traditional weighing methods,and rapid measure the accurate animal’s weight with less affect the animals’ normal activities.There are many types of dynamic weighing algorithms,such as traditional digital filtering,dynamic compensation,and the neural network algorithms.With the development of hardware technology,the computing power of CPU has made great progress,so we can use some middle-calculation algorithm,but still not enough to use,neural network,the large-calculation algorithm.Therefore,the Empirical Mode Decomposition algorithm is selected in this system,as this algorithm run normally in the modern embedded systems and meet the requirements of real-time measurement of animal weight.In order to accurate weighing an animal real-time,the system improve the general EMD algorithm,and adding operations as data preprocessing and early exit.The results show that the final weight error is close to the system error of the data acquisition module,so that the animal weight value can be obtained quickly and accurately,which can promote the research of dynamic weighing for animals.The dynamic weighing system proposed in this paper not only modified the EMD algorithm,but also designed the hardware circuit,implemented the software and tested the system.The hardware part includes collecting circuit,communication circuit and display circuit.The software part includes data acquisition,data preprocessing,improved EMD algorithm,database,data transmission and display.