Title:
COMPUTER NEURAL NETWORK REGULATORY PROCESS CONTROL SYSTEM AND METHOD
Document Type and Number:
WIPO Patent Application WO1992002895
Kind Code:
A3
Abstract:
A computer neural network regulatory process control system and method allows for the elimination of a human operator from real time control of the process. The present invention operates in three modes: training, operation (prediction), and retraining. In the training mode, training input data is produced by the control adjustment made to the process by the human operator. The neural network of the present invention is trained by producing output data using input data for prediction. The output data is compared with the training input data to produce error data, which is used to adjust the weight(s) of the neural network. When the error data is less than a preselected criterion, training has been completed. In the operation mode, the neutral network of the present invention provides output data based upon predictions using the input data. The output data is used to control a state of the process via an actuator. In the retraining mode, retraining data is supplied by monitoring the supplemental actions of the human operator. The retraining data is used by the neural network for adjusting the weight(s) of the neural network.
Inventors:
SKEIRIK RICHARD D (US)
Application Number:
PCT/US1991/005256
Publication Date:
March 19, 1992
Filing Date:
July 25, 1991
Export Citation:
Assignee:
DU PONT (US)
International Classes:
G06N3/04; (IPC1-7): G06F15/80
Other References:
IJCNN INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS vol. 1, 19 June 1989, WASHINGTON,USA pages 209 - 216; WERBOS: 'Backpropagation and neurocontrol : a review and prospectus'
see page 209, left column, line 1 - right column, line 25
see page 210, right column, line 2 - page 213, right column, line 19; figures 1-6
IJCNN INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS vol. 3, 17 June 1990, SAN DIEGO,USA pages 155 - 160; BABA: 'Explicit representation of knowledge acquired from plant historical data using neural network'
see page 155, line 1 - page 160, line 30; figures 1-4
IJCNN INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS vol. 3, 17 June 1990, SAN DIEGO ,USA pages 309 - 314; WANG: 'Self-Adaptive Neural Architectures for Control Applications'
see page 309, line 1 - page 312, line 12; figures 3.1-3.3
ISA PROCEEDINGS. vol. 45, no. 2, 14 October 1990, PITTSBURGH US pages 991 - 1004; SCHNELLE: 'Using neural based process modeling for measurement inference' see the whole document
see page 209, left column, line 1 - right column, line 25
see page 210, right column, line 2 - page 213, right column, line 19; figures 1-6
IJCNN INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS vol. 3, 17 June 1990, SAN DIEGO,USA pages 155 - 160; BABA: 'Explicit representation of knowledge acquired from plant historical data using neural network'
see page 155, line 1 - page 160, line 30; figures 1-4
IJCNN INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS vol. 3, 17 June 1990, SAN DIEGO ,USA pages 309 - 314; WANG: 'Self-Adaptive Neural Architectures for Control Applications'
see page 309, line 1 - page 312, line 12; figures 3.1-3.3
ISA PROCEEDINGS. vol. 45, no. 2, 14 October 1990, PITTSBURGH US pages 991 - 1004; SCHNELLE: 'Using neural based process modeling for measurement inference' see the whole document
Download PDF: