SN 2018hti is a Type I superluminous supernova (SLSN~I) with an absolute $g$-band magnitude of $-22.2$ at maximum brightness, discovered in a metal-poor galaxy at a redshift of 0.0612. We present extensive photometric and spectroscopic observations of this supernova, covering the phases from $sim -35$ days to more than +340 days from the $r$-band maximum. Combining our $BVgri$-band photometry with {it Swift} UVOT optical/ultraviolet photometry, we calculated the peak luminosity as $sim 3.5times10^{44}$ erg s$^{-1}$. Modeling the observed light curve reveals that the luminosity evolution of SN 2018hti can be produced by an ejecta mass of 5.8 $M_odot$ and a magnetar with a magnetic field of $B=1.8times10^{13}$~G having an initial spin period of $P_0=1.8$ ms. Based on such a magnetar-powered scenario and a larger sample, a correlation between the spin of the magnetar and the kinetic energy of the ejecta can be inferred for most SLSNe~I, suggesting a self-consistent scenario. Like for other SLSNe~I, the host galaxy of SN 2018hti is found to be relatively faint ($M_{g} = -17.75$ mag) and of low metallicity ($Z=0.3~Z_odot$), with a star-formation rate of 0.3 $M_odot$ yr$^{-1}$. According to simulation results of single-star evolution, SN 2018hti could originate from a massive, metal-poor star with a zero-age main sequence (ZAMS) mass of 25--40 $M_odot$, or from a less massive rotating star with $M_mathrm{ZAMS} approx 16$--25 $M_odot$. For the case of a binary system, its progenitor could also be a star with $M_mathrm{ZAMS} gtrsim 25$ $M_odot$.