The problem of control for multi-agent motion: General relations