Abstract
Taylor-series integration is a numerical integration technique that computes the Taylor series of state variables using recurrence relations and uses this series to propagate the state in time. A Taylor-series integration reentry integrator is developed and compared with the fifth-order Runge–Kutta–Fehlberg integrator to determine whether Taylor-series integration is faster than traditional integration methods for reentry applications. By comparing the central processing unit times of the integrators, Taylor-series integration is indeed found to be faster for integration without wind and slower with wind, unless the error tolerance is 10−8 or lower. Furthermore, it is found that reducing step sizes to prevent integration over discontinuities is not only needed for Taylor-series integration to obtain maximum accuracy but also for Runge–Kutta–Fehlberg methods. In that case, the Runge–Kutta–Fehlberg integrator does become several times slower than Taylor-series integration.
Original language | English |
---|---|
Pages (from-to) | 2324-2335 |
Journal | Journal of Guidance, Control, and Dynamics: devoted to the technology of dynamics and control |
Volume | 39 |
DOIs | |
Publication status | Published - 2016 |