The invention provides a method for optimizing two-dimensional aeoplotropism sensor network coverage, and belongs to the technical field of sensation. The method comprises the following steps: 1, initializing a sensor network, and calculating the initial coverage rate of a network according to the initial position, the sensation radius, the sector sensation direction angle, the sensation sector field angle and the like of a node; 2, setting the iteration times N to be 1; 3, setting a random disturbance angle for different sensor nodes so as to obtain disturbance angle vectors of the Nth iteration; 4, acquiring the coverage rates of the Nth iteration when a present network rotates on two positive and negative directions of the random disturbance angle vectors; 5, calculating approximate gradient values corresponding to the coverage rate functions, of the present different sensor nodes according to the disturbance angle vectors and the coverage rates of the Nth iteration; 6, updating the sensation direction angles of the different sensor nodes, completing the procedure when N is equal to a preset iteration times K, or else turning back to execute the step 3 when N is equal to N+1. By adopting the method, the optimal rotation angles of different sensor nodes of the two-dimensional aeoplotropism sensor network can be obtained, and the optimization efficiency is high.