：Aiming at the heavy-load parallel robotic manipulators, other than classic sequential 'topology-dimension-structure' parameter design methodology, this paper proposes a new way to optimize dimension and structure parameters 2 simultaneously. To deal with the low-speed but heavy load working environment, a new type of stiffness performance evaluation index is introduced based on kineto-elastic statics (KES) analysis. This stiffness distribution index, mainly focuses on the stiffness distribution among all parts in the manipulator, along with two other dynamic performance indices and some dimensional constraints, formulate a multi-objective optimization model.Then the particle swarm optimization (PSO) algorithm is introduced to search for the optimum parameters. This method is applied on a widely used TRICEPT parallel manipulator, the dynamic performances before and after optimization are compared to verify this effectiveness of this method.