VOLUME 12 NUMBER 1 (January to June 2019)
Philipp. Sci. Lett. 2019 12 (1) 082-091
available online: May 31, 2019
*Corresponding author
Email Address: joe.cruz@icloud.com
Date Received: December 17, 2018
Date Revised: February 28, 2019
Date Accepted: May 03, 2019
ARTICLE
Utilizing data-based artificial intelligence to enable science-based models and dynamic feedback controllers to adapt to disasters
by Jose B. Cruz, Jr.*1
1National Academy of Science and Technology,
Engineering Sciences and Technology Division,Bictuan, Taguig City, Philippines
Engineering Sciences and Technology Division,Bictuan, Taguig City, Philippines
When disasters (such as floods, typhoons, and other natural calamities) occur, dynamic physical systems might change structurally, and the science based models of the systems should change also. If the new behaviors exhibit instability, it would be imperative that the new situations be brought to safe conditions as quickly as possible by means of appropriate feedback control. In these cases, use of artificial intelligence in controllers that do not utilize feedback control would lead to further human fatalities and damages to property.
Selected results from science-based models and control of dynamic physical systems are collected in this paper because the modeling and control methodology for these classes of problems could be implemented automatically. Commercial software systems are widely available for designing the corresponding dynamic controllers.
Major types of disaster situations could be anticipated and each situation could be associated with a predetermined science based model and corresponding dynamic feedback controller. A front-end data-based AI system could be designed to determine what disaster occurred. The identified situation could be compared with a catalogue list of predetermined disasters to decide which predetermined disaster situation is closest. For example, a Bi-directional Associative Memory artificial neural network might be used. The predetermined corresponding dynamic feedback controller would be switched in automatically.
The stabilization challenges posed by the changed behaviors of a cyber physical system after the occurrence of a disaster strongly suggest that mechanisms for science based models and dynamic feedback controllers be imbedded in the system whereby the data based artificial intelligence system would be used to identify and classify the disaster and the feedback controller would automatically adapt to it.
Selected results from science-based models and control of dynamic physical systems are collected in this paper because the modeling and control methodology for these classes of problems could be implemented automatically. Commercial software systems are widely available for designing the corresponding dynamic controllers.
Major types of disaster situations could be anticipated and each situation could be associated with a predetermined science based model and corresponding dynamic feedback controller. A front-end data-based AI system could be designed to determine what disaster occurred. The identified situation could be compared with a catalogue list of predetermined disasters to decide which predetermined disaster situation is closest. For example, a Bi-directional Associative Memory artificial neural network might be used. The predetermined corresponding dynamic feedback controller would be switched in automatically.
The stabilization challenges posed by the changed behaviors of a cyber physical system after the occurrence of a disaster strongly suggest that mechanisms for science based models and dynamic feedback controllers be imbedded in the system whereby the data based artificial intelligence system would be used to identify and classify the disaster and the feedback controller would automatically adapt to it.
© 2024 SciEnggJ
Philippine-American Academy of Science and Engineering