| 摘要 |
A driver condition assessment device that assesses whether the driver of a vehicle is in a condition of decreased consciousness, wherein a main component for the distribution of the acceleration operation condition with respect to the proximity when a preceding vehicle is being followed is analyzed (S16), all of the data is origin-shifted (S18), and a normal driving model of the acceleration operation condition of the driver when the driver is conscious is created (S20). Then, the likelihood of the acceleration operation condition, a likelihood average, a likelihood variance, and a likelihood threshold value are calculated (S22-S26), an assessment is made regarding whether the likelihood of the current driving operations has fallen below the likelihood threshold value (S28), and when it has fallen below that value, the driver is assessed as being in a condition of decreased consciousness. Thus, the assessment is made by calculating a likelihood threshold value on the basis of data when the driver is conscious, so it is possible to prevent an erroneous assessment of decreased consciousness due to characteristic differences in driver operations. |
