Design And Development Of Intelligent Sow Feeding Management System

Pig grain stabilizes the world,while the production and management of sow is an important link in the whole pig industry.In order to improve the production performance and scientific management level of sow,cut down the manual costs,increase the economic efficiency of pig farms,many automated sow feeding systems at home and abroad have been developed in succession to achieve automatic feeding and large-scale management of sows.However,in terms of the existing mature sow feeding system,on the one hand,has a high cost of introduction and purchase,and other hand,it does not fully meet the actual specific needs in variety of places.Therefore,based on the mature experience of the existing system,It is necessary to develop a sow feeding system which is suitable for the local specific conditions.Located in the southwestern part of China,Chongqing has obvious features of high temperatures,hilly and foggy,it is also an important base for pig production of China.In order to meet the demands of the national pig industry development,the technology development,and the Chongqing pig industry,our school has applied and was approved the subject of Chongqing Municipal Science and Technology Commission,we decided to independently develop the "Management System Of Intelligent Sow Feeding(MSOISF)",focusing on the research and development of the sow feeding system which is suitable for the local pig farms in Chongqing.As a member of the project team,the author is mainly responsible for the module of software development.After checking literature in the early stages and conducting field surveys,two main problems are found in the followings: 1)the computer operation level of pig farm manage personnel is generally low,and the previous complex human-computer interaction interface has to some extent limited the promotion and use of the system;2)The automation of existing systems is mainly reflected in aspects of industrial control and Web intelligence,such as automatic distribution and feeding,automatic warning and reminders,etc.However,there is still huge space for research and development on the applications and functions based on artificial intelligence and machine learning.In response to the above problems,this article has carried out the following research work:(1)In the face of the needs of simple and easy-to-use interface,the design of high interaction,easy to operate the front-end interface.This article has introduced a highinteraction design concept,which improves the interactivity and ease of operation of the MSOISF interface from perspectives of navigation bar design,alternate display of multiple tables,and interactive response,so as to achieve a front-end interface which allows pig farm managers easier to understand and accept and are able to quickly grasp the operation method.(2)Introducing the concept of standardized large-scale feeding,conducting demand analysis and framework design for each module of the MSOISF project(such as feeding plan,alarm reminding and processing,production plan,etc.),and has realized the core function module of the system-the feeding plan setting module of automatic induction feeding.(3)Basing on the convolutional neural network algorithm,the intelligent and accurate calculation function of the sow feeding amount has been realized.First of all,through establishing the model to confirm the influencing factors that affect the feeding amount;Secondly,introducing the data set collected by the research into the training model,through continuously adjusting the parameters to obtain the final model;Finally,introducing the model into MSOISF to achieve the effect of intelligent and accurate calculation of sow feeding amount,which has overcome the shortcomings of traditional manual input of sow feeding amount and the inaccurate feeding amount.(4)Basing on the BP neural network algorithm,the automatic assessment of the risk of sows suffering from respiratory syndrome has been realized.First is to establish the BP neural network model,determine the impact factors through research and reference;secondly,using the data collected from the survey to train the model;finally,introducing the trained network into the MSOISF system to achieve the effect of risk assessment of sows with respiratory syndrome.Through the function of this module,a risk pre-judgment reference is provided for the administrators and breeders,which is helpful to inspect the sows in time and take preventive measures and treatment measures in advance,and effectively reduces the farm economy loss.The research results show that,a high-interaction design concept is introduced in this article,through a series of designs such as interface navigation,alternate display of multiple tables,and interactive response,the high interactivity and ease of operation of the MSOISF system interface have been effectively realized;through introducing the convolutional neural network algorithm,the function of the system intelligently and accurately calculating the feeding amount of sows has been realized,through introducing the BP neural network algorithm,the function of systematic risk assessment of sows with respiratory syndrome has been realized,which has made a good attempt for the application of artificial intelligence and machine learning in automatic sow feeding systems,and at the same time provided a good reference for the further development of other intelligent modules in the system.