The human immune system is a highly sophisticated system of biological structures that defends the human body against harmful antigens. In this paper, the concept of biological immune system, idiotypic network theory and clonal selection theory are applied in multi-robot cooperation. Idiotypic network theory governs the stimulation and suppression of an antibody in response to an antigen. Clonal selection theory represents how immune cells mutate and proliferate into memory cells. The methodology presented in this paper is autonomous and de-centralized. In the proposed system, the robots work effectively to complete a task either alone or co-operatively. Specifically, the coordination and communication among various robots are established based on the concept of immune system, idiotypic network theory and clonal selection theory. The developed methodology is successfully implemented through simulation. The experiments show that the simulated robots completed all tasks effectively.