A survey on Hilbert-Huang transform: Evolution, challenges and solutions

•This paper describes the evolution of HHT, focused on its one-dimensional version.•Research limitations and solutions available in the literature are listed.•Some variations of the EMD are compared.•Bibliometric analysis is quantitatively conducted.•Research gaps and suggestions for future work are summarized. Signal processing methods are essential in scientific research, and time-frequency analysis techniques such as Fourier Transform constitute an important progress in data analysis, but may be limited to examining non-linear and non-stationary processes, although there are ways to work around this problem, with Wavelet Transform for example. Recently Hilbert-Huang Transform (HHT) was created, considered by several researchers to be the most appropriate tool to deal with non-linear and non-stationary signals, because unlike the two transforms mentioned, it is not limited by the uncertainty principle and its basis is adaptive. Like other techniques, HHT has some limitations but has great potential and, consequently, the interest of the academic community has grown in recent years, leading the method to be applied in various knowledge areas. In this paper, a literature review that comprises the method since its introduction in 1996 to March 2021 is presented, covering its evolution and some of the solutions proposed over the years for the main problems cited in the literature, which were divided into the five categories: Stop Criterion, Determination of Envelopes, End Effects, Under/Overshoots and Mode Mixing. Some quantitative results are also presented regarding the number of publications by year and type, geographic distribution of research and application areas cited in the papers selected for this survey. It is believed that the exposed in this paper summarizes the evolution of HHT and the studies that involve it, indicating opportunities for research for a tool that has been increasingly powerful.