<p>In this paper, the problem of distributed optimization subject to a convex set is investigated by employing a continuous-time multi-agent system. In the problem, each agent only has access to its own objective function and its own state information, and can only communicate with its neighbors through an unbalanced digraph. Different from most existing works on distributed optimization, we focus on the case where the objective functions are strictly pseudo-convex. To handle this challenge, a continuous-time distributed algorithm is proposed based on the projection operator and the gradient rule. In the proposed algorithm, two consensus strategies are employed. One is used to facilitate the consensus of agents, and the other one is used to estimate the left eigenvector associated with zero eigenvalue of the Laplacian matrix of the unbalanced digraph. Under mild assumptions on the objective functions, we prove that the multi-agent system asymptotically achieves consensus and the consensus state is the solution to the optimization problem. Finally, a simulation example is presented to verify the effectiveness of the theoretical results.</p>