This paper addresses a resource-allocation problem extracted from real life application involving multi-resource operations. A new mixed integer linear programming model is proposed to minimize the weighted completion time while considering resource-related precedence relationships. Then, a hybrid algorithm combining Benders decomposition and Tabu search is developed based on Benders decomposition as the basic framework. This method divides the original problem into a master problem for resource allocation and a subproblem of calculating the completion time of each operation. The convergence is sped up by improving the mathematical model and embedding the Tabu search approach. The experimental results on 300 randomly generated instances show that when solving small-scale problems, the proposed hybrid algorithm can yield satisfactory solutions with an average deviation of 0.86% from optimal ones provided by the commercial CPLEX solver; when solving large-scale problems, the proposed algorithm outperforms the CPLEX solver, the pure Tabu search algorithm, the variable neighborhood search algorithm and the Benders decomposition with embedded genetic algorithm. Compared with the CPLEX, the upper bound and lower bound are improved by 4.74% and 9.62% respectively.