离散控制理论在显示积分中应用研究.pdfVIP

The 12th International Symposium on Structural Engineering APPLYING DISCRETE CONTROL THEORY TO DEVELOP AN EXPLICIT INTEGRATION ALGORITHM WITH UNCONDITIONAL STABLILITY AND CONTROLLABLE NUMERICAL DAMPING FOR REAL-TIME TESTING Cheng Chen School of Engineering, San Francisco State University, San Francisco, CA, U.S.A. Abstract: Integration algorithms are typically utilized in structural dynamics to obtain solution to the temporally discretized equations of motion. Stability is an important issue to be considered when selecting the proper integration algorithm for analysis of structures with a large number of degrees of freedom. Recent development of real-time structural testing brings more challenges to researchers selecting appropriate integration algorithms. An explicit algorithm is more favorable because of its computational efficiency but often constrained by its conditional stability. Moreover, numerical errors often lead to the spurious growth of high-frequency response in the dynamic analysis and the presence of inevitable experimental errors will aggravate this effect in real-time structural testing. It is therefore desirable for an explicit algorithm to not only have unconditional stability but also possess numerical damping to suppress any spurious participation of the high-frequency response while the lower modes can be integrated accurately. This paper presents the development of an unconditionally stable explicit integration algorithm with controllable numerical damping using discrete control theory. Keywords: Intergration algorithm, real-time testing, numerical damping, stability, explicit 1 INTRODUCTION stability is generally preferred over cond