Container in-transit state abnormity discrimination method based on multi-sensor perception

The invention discloses a container in-transit state abnormity discrimination method based on multi-sensor perception. In-transit state data of a container is collected and processed by use of different types of sensors. According to state quantities detected by the different sensors under different transport modes, the influence degrees of the in-transit states of the container in the corresponding modes are endowed with weight values; according to the state quantity given by each sensor, an alarm limit is set, abnormity degrees are set, and quantification processing is performed in terms of the different degrees; the sum of the product of the state quantity values and the weight values of all the sensors is taken as a determination basis for determining different transport states of the container under the condition of diffident transport modes; and a determination standard is given, the in-transit transport state of the container is determined in real time, and an early warning is given in case of an abnormal state. The accuracy and the reliability of a determination result are improved; the determination result of a multi-modal transport condition of the container is more scientific and accurate; and visual numeralization is employed as a determination mode, the classification boundary is clear, and more visual basis references can be given to monitoring decision-makers.