Variational Approach for the Adapted Solution of Backward Stochastic Differential Equations with Locally Lipschitz Diffusion Coefficients

One existence integral condition was obtained for the adapted solution of the general backward stochastic differential equations （BSDEs）. Then by solving the integral constraint condition, and using a limit procedure, a new approach method is proposed and the existence of the solution was proved for the BSDEs if the diffusion coefficients satisfy the locally Lipschitz condition. In the special case the solution was a Brownian bridge. The uniqueness is also considered in the meaning of ＂ F0-integrable equivalent class＂ . The new approach method would give us an efficient way to control the main object instead of the ＂noise＂.