For a class of uncertain systems with only output available and nonlinearities in the presence of unmeasured state，a state observer
based on input signal replacement is firstly established by using the neural networks，then，an output feedback adaptive control
scheme is proposed by employing dynamic surface theory． The controller parameters' range is initially selected by theoretical analysis． It
is guaranteed that all signals in the closed － loop system are semi － global uniformly ultimately bounded． Parameters optimization strategy
is proposed by using genetic algorithm． Then the optimal initial input and controller parameters are chosen，so the problem of input
impulsion is solved and tracking precision is improved．