Method and system for monitoring behavior state of target object

The invention discloses a method and system for monitoring a behavior state of a target object. The method comprises the steps: determining spatial position information on the target object accordingto signal intensity data and barometric information of a position where the target object is located, and determining a pace action type of the target object by utilizing a trained convolutional neural network model according to a triaxial acceleration level of the target object; and determining a current behavior state of the target object according to the spatial position information on the target object and the pace action type. According to the method and the system, disclosed by the invention, the behavior state of the target object can be accurately monitored in real time, and warning can be carried out promptly when the target object presents abnormal behavior states such as approach to dangerous zones, divorcement from operating posts and falling over, so that compared with the existing monitoring methods,