Genetic Algorithm based Location Planning for Object Traking in RFID Enabled Warehouse

Abstract

Warehousing is an operation that aids companies with, systematic and orderly storage of goods on a large scale and making them available conveniently when needed. A warehouse management system which primarily aims to control the movement and storage of materials within a warehouse and process the associated transactions, like shipping, receiving, picking and put away will be a right solution. The recent advancements in chip manufacturing technology are making RFID applicable for a wide range of real time settings. These advancements possess the potential to revolutionize overall supply chain management. In this thesis, the WMS problems have discussed above solves through the hybridization algorithm and cloud technology to reduce the time, cost and connectivity of all warehouses inventory and stock through cloud technologies. The combination of the RFID, Cloud technology and hybridization algorithm for the sorting of racks and locating the packets, forms a new approach in this thesis and termed as Utilitarian model based WMS. In Utilitarian model based WMS, the goods arrive to the warehouse, they are unloaded with the help of a conveyer belt arrangement. Each good/ product possesses a RFID tag attached to it, which contains specific information about it. A RFID reader is placed near the conveyer belt so that it reads the tag data that is carried by each package. This data fed to Matlab VI software, which is run on a computer system. From the computer, data is sent to a transmitter. An unmanned forklift that contains a transceiver is used in shelving the packages to their respective racks. This information is provided to the forklift by means of the transmitter. Thus packages are shelved based on the information received. The arrangement and retrieval of the packages i.e., the location planning of the goods is programmed using Genetic algorithm and hybrid algorithm such as FNNPSO, FNNGA and FNNPSOGA in order to optimize time and cost incurred in warehousing. From the analysis the stipulated newlinevi newlineparameters like fitness function, time of retrieval and fitness value, the efficiency of automatic warehouse system, allocation strategy for database storage including improved FIFO is enhanced, it is seen that faster frequency of moving, easier location availability, better time of retrieval and improved read range is obtained. The analysis showed that the GA algorithm showed excellent discriminatory power and had higher precision, recall, and f-measure values of 0.83%, 0.76% and 0.79% respectively and this algorithm works better than the manual decision making method. newline