Abstract
In the field of cold chain logistics, the key point is the real time control of temperature. Thus the failure of synchronous temperature monitoring is the bottleneck of cold-chain temperature monitoring. Targeting at the real-time features of synchronous temperature monitoring, this paper discusses some issues about RFID technology applied to outlier detection. Through comparing differing feasible RFID data mining methods, along with the requirements of the cold chain temperature monitoring, we put forward the improved QOD (quick outlier detection) algorithm by clustering based on data stream. After that, we prove that QOD algorithm’s performance can be improved after optimization and compared it with several other related methods in accuracy, memory consumption. DOI: http://dx.doi.org/10.5755/j01.eee.19.3.3699
Highlights
Cold chain was introduced by Albert Barrier and O
The technologies are mainly put in practice in the production of some aquatic food, meat and canned food while very rarely applied in the transportation and circulation stages, which leads to its inability to ensure the quality of frozen food in transportation
Among all the factors of cold chain logistics, the temperature is the most direct and easiest one to deal with, so the safety and quality of cold chain logistics should focus on the temperature monitoring
Summary
Cold chain was introduced by Albert Barrier and O. Due to the overwhelming data provided by temperature monitoring of RFID cold chain logistics, we need some effective data stream mining methods to process it. RFID cold chain real-time temperature monitoring aims at finding those temperature points, which are significantly different from others. Other data stream mining methods have remarkable drawbacks in some aspects In this way, outlier mining may gracefully meet the requirements of RFID cold chain real-time temperature monitoring [3]. Statistics can be further divided into methods based on distribution and depth of data [4], [5]. Since RFID cold chain outlier mining algorithm deals with the temperature data which is one-dimension and the real-time temperature data satisfy the normal distribution, we choose the method based on statistics
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
