Main advantage of redundant space manipulator is that it can be configured differently by actuating various joints. Space manipulators are designed to be deployed in harsh surroundings, which in turn would increase the possibility of manipulator failure. Thus, the enhancement of the capabilities of the robots and the failures thereafter are a major concern. This article presents three aspects of hyper-redundant space robot. First, the control strategies based on minimum power criterion with all healthy joints are proposed. Second, hyper-redundancy is put upon to reconfigure the space robot with the failure of locked joints. Third, the optimization of workplace trajectory of the manipulator is obtained by genetic algorithm and a minimum power criterion has been adopted in all three cases. The bond-graph technique is employed for robot modeling with the use of SYMBOLS Shakti software, which generates system equations in C++ code. This code is subsequently converted into a MATLAB program. The result obtained through simulation shows that 37% of the total power consumption was reduced by the actuators through an optimization approach.