||Basic and applied (No. 37145)
||The Science and Technology Development Fund (STDF), Egypt.
||Cairo University Faculty of Engineering
||2020 - 2022
The purpose of this project is to develop a software tool based on artificial intelligence to control tall buildings during earthquakes and avoid damage. This topic is very relevant to the current attitude of Egyptian government in the direction of building the tallest building in the new administrative Capital and other tall buildings in Al-Alamein. The idea of the project is to start from the output of previous STDF project (number 14910). In which the PI and his team developed innovative coupled FE-BE based software tool to analyze tall building under static loading.
In this project the building solution will be extended to account for dynamic loading (earthquake) with damping to simulate reality. The solution will be coupled with semi-active control toolbox after providing diagonal magnitorhological (MR) dampers. The places of such dampers will be computed based on optimization scheme.
In order to speed up the process, the solution techniques is going to be replaced by using artificial neural networks (ANN) via GPU computing. Hence, practical applications could be solved in mater of few seconds. The outcome of the project is a software tool to compute the ANN for any tall building (with or without the dampers). The training data for this ANN is obtained from developed software that analyzes buildings under earthquake loadings. This software could be installed and connected to dampers on building site and activated when earthquake hits the area and consequently avoid damage and save lives and money. The project will have one researcher plus the PI and Co-PI. The PI and Co-PI have collaborated several times together and the full team with the young researcher already have a synergetic effect together.
The objectives of this project are:
The general objective of this research is to design semi-active control laws to mitigate the vibrations in adaptronic systems, especially in civil engineering structures equipped with adaptive dampers. An exhaustive theoretical study of adaptive dampers is to be performed in addition to developing algorithms capable of intelligently controlling the dampers. Thus, the goal is to develop models that are suitable for control design, whether the models are done numerically or through trained artificial intelligence. New semi-active control strategies will be developed in order to effectively deal with the complexity of hysteretic nonlinearities, parametric uncertainties, measurement limitations and unknown disturbances, and consequently to achieve the robust performance. The formulation of semi-active control laws will be done by introducing, from other application fields, the useful control methodologies into the field of vibration mitigation in adaptronic systems (some of them for the first time) or by making improvements of other control methodologies already applied in this field. The performance of the semi-active control laws will be validated by means of numerical simulations and machine learning simulations. The project’s brief objectives are:
1- To develop artificial neural network based active control software tool to be installed on the relevant building site to automatically adjust, during earthquake, the properties of dampers, consequently avoid damage, and save lives and money.
2- To develop a software tool that analyzes tall buildings under earthquake loadings to generate the required training data for objective no 1.
Building and construction greatly flourished in the last decade and took its place as our important major driver of economic growth in Egypt. In 2016, it provided an increase by 10.3%, a significant uptick after growing at an average of 5.3% in the preceding four years. This growth trend has largely been maintained into 2017, growing at a rate of 7.3%. Moreover, it greatly helps in solving the unemployment issue by providing several job opportunities. Furthermore, it encourages investors to conquer new areas and invest in housing projects (Data is obtained from the ministry of planning).
Dynamic load analysis and optimization are essential needs especially when Egypt is entering a new era in the new Capital. A high-rise building might be load by an oscillating wind load or earthquakes, regardless the cause of vibration, one should consider these issues in the design process by upgrading the new buildings with a new tool to minimize vibration or displacements.
The traditional way to bypass oscillating loads might not be efficient due to the design complexity; performance-based building design approach is needed to fulfill our design criteria.
It is no longer a choice for the construction companies to use performance-based design for buildings because it is now an official demand as Egypt is now interested in building high-rise buildings.
A performance-based building is a construction built to meet an expected or measurable requirements performance. This building is upgraded by a controller that is trained (from previous data) using neural network to deal with undesired disturbances and minimize or maximize certain objective function performance. Using traditional methods to deal with such disturbances would be very expensive, on the other hand, A performance-based building would adapt to any new inputs and learn from previous data and give an accurate record for the earthquakes without any additional costs. A building with these capabilities would be more effective to avoid losses and would give a better understanding to the earthquakes and wind dynamics on the building.
This project simply offers a tool to control high-rise buildings and to achieve the performance-based requirements; this greatly fits to the county’s ten years plan 2030.