No abstract

No abstract

Real-time hybrid simulation (RTHS) is intended to serve as a technique able to conduct experiments when the behavior of the plant is not well understood, i.e., when deep uncertainties are present in the physical specimen. By combining strategies from robust control and adaptive control, this paper develops an adaptive sliding mode control (ASMC) system for uncertain control plants. The ASMC consists of a bounded-gain forgetting least-squares estimator and a sliding mode controller, aimed at estimating parameters of the control plant and eliminating the negative effects of estimation errors, respectively. The ASMC is evaluated by applying it to the benchmark control problem in RTHS, where the fifth-order control plant is reduced to a second-order control plant to facilitate the control system's design and execution. High performance and robustness are achieved with the adoption of ASMC. The results demonstrate that an effective ASMC can be designed based on a significantly simplified control plant, making it a potent control system for RTHS. K E Y W O R D Sadaptive control, model uncertainties, real-time hybrid simulation, robust control, sliding mode control | INTRODUCTIONReal-time hybrid simulation (RTHS) is a cyber-physical technique to conduct experimental studies of large-scale and complex structures subject to dynamic loading. 1 In RTHS, the structure under investigation is divided into the numerical substructure-computational model simulated in a computer and the physical substructure-experimental specimen(s). The numerical substructure should contain the majority or larger part of the structure. It is typically assumed to be well understood and can be numerically modeled precisely, while the rest of structure, normally individual elements(s) or device(s) inside of the structure, e.g., a damper, is treated as a physical substructure and is

Real-time hybrid simulation (RTHS) is an enabling technology that has transformed engineering experimentation and helped researchers expand modeling capabilities. However, breakthroughs are necessary to expand the range of hybrid simulation methods and, thus, enable experiments with loading conditions representing multiple hazards. This paper discusses the development of a new thermomechanical RTHS framework and a systematic approach to determining RTHS control requirements. First, the framework is established using a representative finite element model of a layered structural system subjected to thermal loading. A complete two-layer system model serves as the reference system, and it is then partitioned into a numerical layer and an experimental layer that share interface conditions. Next, a thermal actuator is introduced to impose dynamic thermal loading on the experimental subsystem, serving as a transfer system. Finally, control and performance metrics are defined to evaluate the realization of interface boundary conditions and map this to the RTHS execution. Through an illustrative example considering the influence of temperature on a lunar habitat, we demonstrate how to establish controller requirements for RTHS and demonstrate that this approach can be used to conduct RTHS on structures with thermomechanical loading.

Currently, the lack of (1) a sufficiently integrated, adaptive, and reflective framework to ensure the safety, integrity, and coordinated evolution of a real-time hybrid simulation (RTHS) as it runs, and (2) the ability to articulate and gauge suitable measures of the performance and integrity of an experiment, both as it runs and post-hoc, have prevented researchers from tackling a wide range of complex research problems of vital national interest. To address these limitations of the current state-of-the-art, we propose a framework named Reflective Framework for Performance Management (REFORM) of real-time hybrid simulation. REFORM will support the execution of more complex RTHS experiments than can be conducted today, and will allow them to be configured rapidly, performed safely, and analyzed thoroughly. This study provides a description of the building blocks associated with the first phase of this development (REFORM-I). REFORM-I is verified and demonstrated through application to an expanded version of the benchmark control problem for real-time hybrid simulation.