Particle crowd algorithmic rule is a mayor examination hotspot in the authentic optimization algorithmic rule respond. Based on the PSO algorithmic rule to make optimal the RBFNN example, an amended order of nonlinear adaptable laziness power supported on the contest of population variegation is intended to extend the fixedness of population unlikeness performance and hunt capabilities to preclude the algorithmic rule from dripping into a topical extreme point prematurely, thereby further improving the prophecy correctness. Simulation experience shows that the amended PSO-RBFNN standard has open advantageous in the fixedness and sharp convergency of the prognosis proceed. In fashion to reprove the justness of reverse kinematics of robots with composite make and supercilious degrees of liberty, an amended adaptative suffix abound optimization (IAPSO) is spoken. First, the motoric equality of the 6-DOF strength-example avaricious robot design is established by the amended DH (Denavit-Hartenberg) argument course; second, on the base of the existent morsel abound algorithmic rule, the population Manhattan ceremoniousness is interested to lead the maneuver condition of the population in aqiqiy measure. And bound the adaptative lore substitute accordingly to the dissimilar maneuver possession and then adopt distinct site and hurry update modes; lastly, the fitness province with handicap substitute is present to trial the honest-prick and extended course transposition of the robot mold, and the delusion is not joint product major than 0.005 rad. The feint inference shows that the established kinematics shape is chasten, and the amended algorithmic program captures into recital the nicety, uniqueness, and velocity of the inverted resolution of the existent PSO algorithmic program, as well as higher deliverance truths. We conduct an experiment on the Brazilian jiu-jitsu. The results have clearly shown the advantage of our method.