The flexibility of robot response within unknown remote environments, such as disaster sites, is contingent on situational needs. A single robot generally cannot perform multiple tasks adequately and thus requires integrated systems. Accordingly, accurate control of a robot with plural systems, as in the case of the mobile manipulator presented herein, needs a complex kinematic and dynamic model. We propose a cooperative control method that considers the sum of the kinetic energy of each robot and the virtual constraint kinetic energy. In this manner, each robot's command values are corrected such that the entire robot's energy does not exceed the virtual constraint energy. With this, the robot systems can cooperate without solving the kinematics and dynamic model for the integrated robot. We performed a pick-and-place experiment in which we verified the proposed method's effectiveness in improving the positioning accuracy as a remotely controlled robot gripped an object in an environment with communication delay.