بهینه سازی ساختاری سایبری فیزیکی
Abstract
1. Introduction
2. Real-time hybrid optimization (RTHO)
3. Experimental setups
4. Results
5. Conclusions and recommendations
Acknowledgements
References
Abstract
Traditionally, structural optimization is a numerical process; candidate designs are created and evaluated through numerical simulation (e.g., finite element analysis). However, when dealing with complex structures that are difficult to model numerically, large errors could exist between the numerical model and the physical structure. In this case, the optimization is less meaningful because the optimal results are associated with the numerical model instead of the physical structure. Experiments can be included in the optimization algorithm to represent complex structures or components. However, the time and cost limitations are prohibitive when iteratively constructing and evaluating complete structural systems. Real-time hybrid simulation (RTHS) is an efficient and cost-effective experimental tool that combines numerical simulation with experimental testing to capture the total structural performance. This paper proposes a framework for real-time hybrid optimization (RTHO); RTHS is used to evaluate the performance of candidate designs within the optimization process. The framework creates a cyber-physical optimization environment using RTHS, a modern experimental technique with roots in earthquake engineering. This paper outlines the framework for RTHO with accompanying proof-ofconcept studies. In a preliminary study, the base isolation design of a two-story building was optimized for seismic protection. RTHO was further validated for the optimal selection of multiple semi-active control law parameters for an MR damper installed in the isolation layer of a five-story base-isolated building. Both cases used RTHS to evaluate the candidate designs and particle swarm optimization (PSO) to drive the optimization. RTHO is well-suited to evaluate nonlinear experimental substructures, in particular those that do not undergo permanent damage such as structural control devices. Structural damage, if of interest, can be modeled through the numerical component. This paper proposes and demonstrates the integration of state-of-the-art optimization algorithms with state-of-the-art experimental methods – a cyber-physical approach to structural optimization.
Introduction
In the last two decades, design trends for civil infrastructure have shifted from prescriptive code procedures to performance-based methods. The structural engineers of tomorrow will be asked to produce lighter, taller, and more cost-effective designs to meet performance demands under natural and human-made hazards. Sustainable solutions that consider long-term costs and benefits will require new approaches to design and optimization. Structural optimization enables engineers to minimize user-specified objectives (e.g., material use) while ensuring strength and serviceability requirements constraints are met (e.g., requirements for drift and acceleration). Structural optimization includes size optimization, shape optimization, and topology optimization [1]. Size optimization focuses on optimizing the cross-section of the discrete structural members such as beams and columns, or thickness of continuous material such as panels and slabs. Shape optimization allows the location of nodes and connections to vary. Topology optimization uses the distribution of material and structural connectivity to find the optimal layout of the structure for a given shape.