ارزيابي عملكرد موجودي مبتني بر پيش‌بيني تقاضا با استفاده از مدل شبكۀ عصبي MLP

Inventory performance evaluation based on demand forecast with Neural Networks (MLP) approach

ﻣﺪﻳﺮﻳﺖ ﺻﺤﻴﺢ و ﻛﻨﺘﺮل ﺑﻬﺘﺮ ﻣﻮﺟﻮدي اﻗﻼم ﻓﺮوﺷﮕﺎه ﻣﻮاد ﻏﺬاﻳﻲ، ﻳﻜﻲ از ﺿﺮوريﺗﺮﻳﻦ و ﻣﻬﻢﺗﺮﻳﻦ اﻫﺪاف ﻣﺪﻳﺮان ﻓﺮوﺷﮕﺎهﻫﺎي ﻣﻮاد ﻏﺬاﻳﻲ ﻣﻲﺑﺎﺷﺪ. ﻫﺮ ﭼﻪ ﻋﻤﻠﻜﺮد ﻣﻮﺟﻮدي ﻓﺮوﺷﮕﺎه ﺑﻬﺒﻮد ﻳﺎﺑﺪ، اﻓﺰاﻳﺶ ﺳﻄﺢ ﺧﺪﻣﺖ ﺑﻪ ﻣﺸﺘﺮي و ﻛﺎﻫﺶ روزﻫﺎي ﻣﻮاﺟﻪ ﺑﺎ ﻛﻤﺒﻮد را در ﭘﻲ ﺧﻮاﻫﻴﻢ داﺷﺖ. ﻫﺪف اﻳﻦ ﻣﻘﺎﻟﻪ اراﺋﻪ ﻳﻚ ﻣﺪل ﭘﻴﺶﺑﻴﻨﻲ ﺑﺮاي ﺗﻘﺎﺿﺎي ﻓﺮآوردهﻫﺎي ﮔﻮﺷﺘﻲ ﻓﺮوﺷﮕﺎه زﻧﺠﻴﺮهاي اﺗﻜﺎي ﮔﺮﮔﺎن، ﺑﻪﻣﻨﻈﻮر ﺑﻬﺒﻮد ﻋﻤﻠﻜﺮد ﻣﻮﺟﻮدي ﻣﻲﺑﺎﺷﺪ. در اﻳﻦ ﭘﮋوﻫﺶ از ﻣﺪل ANNmlp ﺑﺮاي ﭘﻴﺶﺑﻴﻨﻲ ﺗﻘﺎﺿﺎي ﻓﺮآوردة ﮔﻮﺷﺘﻲ اﻳﻦ ﻓﺮوﺷﮕﺎه اﺳﺘﻔﺎدهﺷﺪه اﺳﺖ و ﻫﻤﭽﻨﻴﻦ ﺑﺮاي درك ﻣﻴﺰان دﻗﺖ ﭘﻴﺶﺑﻴﻨﻲ، ﺑﺎ ﻣﺪلﻫﺎي ARIMA و ﻣﻴﺎﻧﮕﻴﻦ ﻣﺘﺤﺮك 14 روزه ﻣﻘﺎﻳﺴﻪ ﺷﺪه اﺳﺖ. ﺑﺮاي اﻳﻦ ﻣﻨﻈﻮر، از ﻛﺪ ﻧﻮﻳﺴﻲ اﻳﻦ ﻣﺪل در ﻧﺮماﻓﺰار ﻣﺘﻠﺐ و دادهﻫﺎي ﺳﺮي زﻣﺎﻧﻲ ﺗﻘﺎﺿﺎي ﻓﺮآوردهﻫﺎي گوشتي از ابتداي سال 1392 تا هفتۀ 12 سال 1395 كه به‌صورت هفتگي دريافت گرديد، استفاده شد. نتايج تحقيق نشان داد كه مدل 1-8-5ANN بهترين مدل براي پيش‌بيني تقاضاي اين محصول مي‌باشد. مدل پيش‌بيني ارائه‌شده با سياست كنترل دوره­اي سطح موجودي، منجر به كاهش روزهاي مواجه با كمبود و افزايش سطح خدمت به مشتري شده است.

Proper management and better control of inventory of food items are one of the most essential and important objectives of food store managers. The store inventory management to improve performance, increase customer service and reduce the deficit in the following days will be. The purpose of this paper is to provide a prediction model for demanding meat products from Gorgan ETKA chain stores in order to improve inventory performance. In this study, the ANNmlp model has been used to predict the meat market demand of this store and is also compared with ARIMA and 14-day moving average to understand the accuracy of prediction. For this purpose, the code coding of this model was used in MATLAB software and time series data for meat products demand from the beginning of 1392 to the 12th week of 1395, which was received weekly. The results of the research showed that ANN5-8-1 model is the best model for predicting the demand for this product. The prediction model provided by the inventory control period has led to a reduction in the number of days facing the shortage and increased customer service levels.