Integral maximisation algorithm. Comparison with standard methods and application to adaptive optical systems with statistical backlash

Abstract

One of the fundamental challenges of engineering is establishing the optimal parameters in problems with multiple local solutions. A large number of optimisation algorithms have been developed for this task. This paper proposes a new optimisation algorithm that is effective in systems with hysteresis and non-linearities. Tests are performed with standard optimisation functions and with a complex optical model. For each test, the results of several known optimisation algorithms are compared with the proposed one. All optimisation algorithms used in the tests are then implemented in an actual thermal diffusivity measuring instrument which uses an optical technique. The optical alignment process of this real-world instrument is carried out by various optimisation algorithms and the dispersions in the measured values are compared. It is concluded that the proposed algorithm presents high robustness and the fastest convergence to the global optimum.