Discrete manufacturing industrial data characterization method based on deep reinforcement learning

• Main Authors: WANG CONG, LIU XIN, YANG HAIGEN, LIN DONGHUANG, GE YAN, DAI ERHAN, ZENG FANYU
• Format: Patent
• Language: Chinese; English
• Published: 16.05.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses a discrete manufacturing industrial data characterization method based on deep reinforcement learning, and the method comprises the steps: collecting discrete manufacturing industrial data, and constructing a space-time database; the method comprises the following steps: dividing discrete manufacturing industrial data into discrete features and continuous features, constructing a data coupling coding network, converting coding vectors in the coding network into representation vectors, and constructing a data representation model; the discrimination degree of data categories is quantitatively represented through clustering evaluation indexes; clustering evaluation indexes of different dimensions are weighted to serve as dynamic rewards, a deep reinforcement learning model is constructed, and neural network parameters of deep reinforcement learning are updated through the interaction relation between the representation model and a discrete manufacturing decision analysis system. Accordin