Research Of Near-infrared Spectrometer Quick Quality Inspection Method Of Feed And Its Production Logistics Traceability System

China is an important feed production country in the world,but in the process of development,the situation of the large but not strong and large and inferior of feed industry has not been fundamentally reversed.The main reasons are that the technical level is not advanced enough,the management concept is relatively backward,and the lack of quality and safety guarantees.The production logistics tracing is an important method to optimize the feed supply chain,mainly including information construction and information acquisition.Near-infrared spectroscopy detection technology has outstanding characteristics such as fast speed,high efficiency,non-destructive and pollution-free,etc.,and has become an important means of obtaining quality inspection information in feed production logistics.Combining feed quality traceability with advanced near-infrared spectroscopy detection technology not only enables feed enterprises to supervise production logistics and improve their competitiveness,but also has important significance for the downstream supply chain aquaculture industry and even national food safety.This paper focuses on feed quality inspection,researches the near-infrared spectroscopy rapid quality inspection technology and its application in the process of feed production and distribution,and designs feed production logistics and quality traceability information software.The main research work of this paper is as follows:(1)In this paper,the parallel design of the MSC,SNV,SG preprocessing algorithm and partial least squares regression modeling algorithm(PLS)in the commonly used near infrared spectroscopy detection technology in the process of feed production logistics is selected,and the production materials in the process of feed production logistics are selected Data,considering the generality of the data at the same time,set up random data of different sizes,and conduct a comparative experiment with the traditional serial algorithm,proving that the above parallelized algorithm can well improve the convenience and speed of production logistics quality inspection.(2)This paper proposes a quality inspection model(FAST-PLS-KELM)based on parallelized PLS(FAST-PLS)and Kernel Extreme Learning Machine(KELM).This model uses PLS to select the main component as the input of KELM for feed ingredients.The classification is divided into two categories and multiple classifications.The results show that it has leading accuracy and speed compared with four commonly used near-infrared spectral classification algorithms.(3)Based on advanced feed production logistics quality inspection technology,this paper proposes and designs a feed-oriented production logistics quality traceability system and displays the interface.It mainly implements the production logistics link involving feed raw materials,production and storage and supplements feed delivery to generate a traceability code link to realize the information construction of feed quality traceability.