黏度學習暨預測系統

植物性飲品加工機具與製程精進研發計畫

陳健南；曹凱傑；Kumar Shivam；蕭仁忠；陳哲堅

一種黏度學習系統，具有一桶槽供容設混合物且設有一用以測量混合物黏度之量測單元，桶槽與至少一驅動件相連接，驅動件用以攪拌混合物，而量測單元於攪拌過程中偵測混合物的一黏度輸出變量資訊，一伺服器儲存有預先量測的混合物之濃度資訊，一第一控制單輸入混合物之一第一變量資訊並於攪拌過程中偵測混合物的一第一過程變量資訊，一第二控制單元輸入驅動件之一第二變量資訊並於攪拌過程中偵測驅動件的一第二過程變量資訊，藉由將第一過程變量資訊、第二過程變量資訊及黏度輸出變量傳送儲存至伺服器中，據以建置一用於預測混合物黏度之模型。

A viscosity learning system has a bucket for containing a mixture and a measuring unit for measuring the viscosity of the mixture. The bucket is connected to at least one driving member. The driving member is used to stir the mixture, and the measuring unit is used for stirring the mixture. During the process, a viscosity output variable information of the mixture is detected, a server stores pre-measured concentration information of the mixture, a first control unit inputs a first variable information of the mixture and detects a first variable information of the mixture during the stirring process. A process variable information, a second control unit inputs a second variable information of the driver and detects a second process variable information of the driver during the stirring process, by combining the first process variable information and the second process variable information And the viscosity output variables are sent and stored in the server to build a model for predicting the viscosity of the mixture.