How to reflect some degree of incentive and take into account fairness is one of the key problem of performance allocation. In this paper, all assessment objects are classified based on the individual difference, by introducing conversion function and desirability function regarding the control parameters to reflect the degree of incentive and basic workload of all assessment objects as variables, a multi-objective optimization model is formulated  with the objectives of maximizing the total desirability and the tradeoff of all assessment objects' desirability. Furthermore,  by means of the epsilon-constraint scalarization method, the existence is proved for weakly efficient solution. As its application, the optimal performance allocation of is studied.