In order to solve the problem of the density measurement at high temperature and high pressure (HTHP)

particularly in a supercritical state

a density measurement method at HTHP was designed and a density prediction method based on least square support vector machine (LSSVM) was proposed. This experimental method was used to measure the density at different (T

p) points firstly. Then

by learning from the experimental data

the density prediction model based on LSSVM (DP-LSSVM) could be established

and the continuous density values in T-p plane could be predicted by DP-LSSVM

especially in supercritical region where the density was difficult to measure by experimental means. Hexamethyldisiloxane was taken for example. By using the proposed method

the density function on T (600-850K) and p (1.3-2.25MPa) was obtained

and density in this area could be calculated. By comparing the calculation results with the published data

it is found that the relative error is only 2.4%

the effectiveness of the method is proved.