The article is devoted to the questions of the optimization and robustization of the control of surface-to-air and air-to-air of an automatic maneuverable aerial vehicle (AMAV). AMAV is a unique and rapidly developing type of the aviation and rocket technology designed for autonomous withdrawal and rapprochement with aerial targets having on-board homing systems. In the first part of the article, the problem of parametric synthesis of the AMAV optimal homing system is considered according to the accuracy and energy consumption criteria of the guidance. The two-criterion optimization of the homing system is carried out in the Matlab environment using a genetic algorithm in the conditions of possible three typical maneuvers of an aerial target: a slide, a dive and a dead loop. The second part of the article analyzes the problem of synthesizing robust control algorithms of the AMAV attack angle. Two algorithms for AMAV robust tracking for programmed change in the attack angle are developed: the first implements the control method in sliding modes, and the second deals with the adaptive backstepping control method. The developed algorithms for controlling the AMAV attack angle in the Matlab environment was approbated using a computer.