In order to effectively solve the inverse kinematics problem of redundant manipulators, an inverse kinematics solution based on improved fruit fly optimization algorithm was proposed. The improved algorithm adopted a linear candidate solution generation mechanism, which overcame the shortcomings that the fruit fly optimization algorithm could not search negative space and could not search uniformly in the specified area. Through the construction of hybrid learning olfactory search strategy, the global exploration and local exploitation of the algorithm were effectively enhanced and reasonably balanced. In addition, through the introduction of the real-time visual updating strategy, the search efficiency of the algorithm was improved, and the convergence rate was also effectively accelerated. Taking the inverse kinematics solution of a 7-DOF redundant manipulator as an example, the results showed that the proposed algorithm was superior to the comparative algorithms in terms of convergence rate, convergence accuracy and results stability, which indicated that the method can be used to effectively solve the inverse kinematics problem of redundant manipulators.