Fuzzy-Type Nonlinear Systems and Construction of Lyapunov Functions for te Control of Autonomous Robot

1.The report presents a pattern classification for systems with arbitrary nonlinear feedbacks.Classification is carried using the robust nonlinear -perturbation technipue，其中非linearities are regarded as the non-linear perturbation of linear stable systems. theresults show that the autonomous robot system can keep stable if nonlinearities lie in therobust-stable range.2. We have proposed an algorithm that solves fuzyy relation equationsSup / min or Inf / max composite. The solution is useful to establish fuzzy inference.3. The geneticalgorithm(GA) requires setting a fitness function and evaluat-ing it to solve the probrem. then，for the autonomous acquisition problem of game策略，这并不easy to apply GA因为健身功能不能被固定在可改变的步骤game. The report gives that The proposed co-evolution approach is useful to The problem.4“antecedent” and "consequent" conditions of fuzzy inference systems are correlated with inputand output of conventional nonlinear feedbacks，respectively. The robust perturbation technique shows that if The nonlinearities are within Therobust perturbation limit，混乱控制系统已被verified to the stability theorm of Lyapunov. the报告has investigated the stability of fuzzy contrl systems using thetechnique。

1.The report presents a pattern classification for systems with arbitrary nonlinear feedbacks. Classification is carried using the robust nonlineear-perturbation technipue, in which nonlinearities are regarded as the non-linear perturbation of linear stable systems. The results show that the autonomous robot system can keep stable if nonlinearities lie in the robust-stable range.2. We have proposed an algorithm that solves fuzyy relation equations given by Sup・min or Inf・max composite. The solution is useful to establish fuzzy inference.3. The genetic algorithm(GA) requires setting a fitness function and evaluat-ing it to solve the probrem. then, for the autonomous acquisition problem of game strategy, it is not easy to apply GA because the fitness function can not be fixed in changeable situation of game. The report gives that the proposed co-evolution approach is useful to the problem.4. The "antecedent" and "consequent" conditions of fuzzy inference systems are well correlated with input and output of conventional nonlinear feedbacks, respectively. The robust perturbation technique shows that if the nonlinearities are within the robust perturbation limit, the fuzzy control system is verified to be stable according to the stability theorm of Lyapunov. The report has investigated the stability of fuzzy contrl systems using thetechnique.