The present chapter proposes a departmental store automation system based on Radio Frequency Identification (RFID) technology and Particle Swarm Optimization (PSO) algorithm. The items in the departmental store spanned over different sections and in multiple floors. Items are tagged with passive RFID tags. Each tag contains unique identification number for the item. The entire floor is divided into number of zones depending on different types of item that are placed in their respective racks. Each of the zones is placed with one RFID reader, which constantly monitors the items in their zone and periodically sends that information to the application. The problem of systematic and periodic monitoring of the store is addressed in this application so that the locations, distributions and demands of every item in the store can be invigilated with intelligence. The proposed application is successfully demonstrated on a simulated case study.
Top1. Introduction
In this chapter, an algorithm has been suggested, based on Particle Swarm Optimization (PSO), used for tracking objects in a departmental store. Consider a departmental store with different types of items. Those items enter into the store and are placed in proper racks. A passive RFID tag has been tagged with each item and each item receives a unique identity number. The readers are placed in the strategic positions in the store to automate the identification process and receive the information of the items present in its communication range. Here, the submissive tags have been used for minimizing the cost of the system. The radio range of the reader is 50 meters. Hence, the requirement of number of readers to track the items comprised over entire departmental store. The readers store information of the items in their regions and transmit that information periodically to the application for taking appropriate decisions.
The exact information about the number of items present in a store, their locations and the distributions are provided by RFID systems. On the other hand, the devised algorithm deals with information management, improved services and information related to demand and use of items in the store.
Swarm Intelligence (SI) is one of the techniques in the field of Artificial Intelligence (AI) that includes the study of collaborative behavior in decentralized systems. Such systems are developed by a population of individual entities collaborating locally with each other and with their working environment. In such systems no centralized control is available to detect the behavior of the individuals. Local communications among the individual objects often lead to invention of a global pattern with successive iterations. SI refers to the problem solving techniques discovered from the interoperations between the individuals and the computational swarm intelligence, which refers to the algorithmic modeling of such activities and behaviors. These algorithmic models could be well adjusted in changing and dynamic environments and are greatly adjustable and robust in nature. The last decade has observed a rapid and vast growing research activities in SI, especially on the most popular and effective SI paradigm known as Particle Swarm Intelligence.
PSO is a stochastic, population-based optimization algorithm introduced in 1995 by Eberhart and Kennedy (1995). Many variations of PSO have been reported in the literature with time. PSO has been applied in many fields to solve different classes of problems like single and multi objective optimization, constrained satisfaction etc.
This chapter presents design, development and implementation of a RFID based departmental store to periodically monitor the items and to identify the most and least demanded item list using PSO algorithm that in turn would be useful for theft detection and stock updates for that store.