The Hilbert transform is useful in calculating instantaneous attributes of a time series, especially the amplitude and frequency. It is a simple and useful algorithm for instantaneous frequency extraction of a signal. For more information I refer you to help of Matlab with "hilbert" keyword.
Furthermore the Hilbert transform is useful for envelope detection of a signal. EMD (Empirical Mode Decomposition) is also a good example of hilbert transform usage.
HI. The Hilbert transform has a good performance for signals with cycles marked and series with outliers. They are widely used in feature extraction, to process time series, in diagnosis of vibratio, signal processing. In the area of energy is used for transient analysis to determine frequencies, as support for the design of filters.
In my PhD dissertation, I used the Hilbert Space vector formulation (Hilbert Transform) to show the general relationship of it to the Digital Fourier Transform.
More info: http://en.wikipedia.org/wiki/Hilbert_space