In the current power grid, the single channel transmission power accounts for a relatively high proportion, which increases the probability of large power disturbance. In view of the problem of insufficient simulation accuracy of frequency characteristics of power system. The Fibonacci tree optimization (FTO) algorithm, which has the characteristics of global and local optimization, is introduced to solve the problem of the overall identification of the speed control system in a regional power grid. At the cost of the multi-peak optimization of the algorithm, the global optimization ability and convergence rate of FTO are improved by limiting the structure depth of the tree and adjusting the global search strategy. The adaptability of the algorithm to parameter identification is improved and the Fibonacci tree identification algorithm is formed. Based on the actual measured frequency response curve, the algorithm is used to identify the parameters of the turbine-governor system of the actual regional power grid. And the comparison and analysis of the multiple algorithms using for this problem are carried out. The simulation results show that the Fibonacci tree identification algorithm has a good performance in this field.