<p>This study introduces a novel neural estimation method ensuring prescribed performance enabling the online identification of uncertain chaotic plants through which the transient-time and state residuals separately adjusted. The Lyapunov stability concept is applied for guaranteeing the stability in adaptive real-time system identification and control process. Outstanding from precedent researches, both identified scheme and training procedure along with control laws are initiatively designed as to permit separately adjusting the transient-time value from erroneous residual one. Then the real-time estimation process is to be satisfactorily investigated, in which the uncertain parts are parameterized using neural-based learning models, whose parameterized procedure helps efficiently solving more complicated problems, including adaptive observer and advanced adaptive neural-based control approaches. Here hyper-chaotic plants are comprehensively tested as to confirm the superiority and adaptability of proposed algorithm in comparison with other recently published advanced estimation approaches.</p>