隨著無線射頻辨識技術(RFID)的科技進步,越來越多的企業導入RFID系統,要能正常發揮RFID系統的水準就必須做好RFID網路規劃。RFID網路規劃是RFID系統中的核心環節,已有不少文獻在討論RFID網路規劃問題,本研究參照以前學者文獻所提到的目標式與限制式,建構出一多目標數學模式。為了取得最佳網路規劃,在許多相關論文中,使用者需先給定讀取器數量來透過演算法求解,而本研究導入變動長度之多目標基因演算法,對整個模式求解,此演算法會自動搜尋最佳讀取器的數量與位置,並設計一田口實驗設計取得演算法之最佳參數。本研究也設計兩個使用情境來驗證演算法的結果,透過使用情境的結果,可以證明本研究所提出的方法可以有效的解決使用者在無線射頻網路規劃上的問題,也可以提供多組網路佈置解供使用者參考。
With the advance of RFID technology, a lot of enterprises import RFID system. To be able to function properly RFID system performance have to do RFID network planning. RFID network planning is the core-value of RFID system. There are a lot of literature discussing the RFID network planning problem. Our research construct a multi-objective mathematical model which are reference the previous scholars literature’s objective function and constraints. In many relevant paper, the user need to give the number of reader to the algorithm to solve. Our research want to use the variable-length genetic algorithm to solve whole model, this algorithm will automatically search for the best reader number and position. Our research using a Taguchi experimental design to get the optimal algorithm’s parameters. Our research designed two scenarios to validate the results of the algorithm, through the results of the scenarios that can prove our proposed method can effectively solve the RFID network planning problem and provide multiple solution for user.