基于fuzzyTECH智能小车神经模糊控制器的实现

Implementation of Neuro-Fuzzy controller for smartcar based on fuzzyTECH

【目的】设计智能小车神经模糊控制器,并对其参数进行优化。【方法】基于模糊控制理论,利用神经网络自学习的能力,通过fuzzyTECH平台设计并调试智能小车神经模糊控制器。【结果】结合模糊控制器强大的推理能力与神经网络自学习能力,解决了传统PID控制器控制参数固定,难以适应多种路面状况的难题,并克服了传统模糊控制器后期参数优化难的不足。物理验证试验表明,基于该算法的控制器具有较高的稳定性、适应性与实时性。【结论】所设计的控制器具有较强的实用性,可以满足智能车控制器的要求。

【Objective】 The purpose of this paper was to design a Neuro-Fuzzy controller for a smartcar and optimize its parameters.【Method】 The Neuro-Fuzzy controller for a smartcar was designed and debugged through fuzzyTECH,based on fuzzy control theory and the self-learning ability in neural network.【Result】 This work combined the benefits of the strong reasoning ability in fuzzy controller and self-learning ability in neural network.It also solved the problem that a traditional PID controller with fixed parameters was difficult to adapt to different situations and overcome the difficulty in parameters of traditional fuzzy controller for optimizing in the later stage.Physical experiments showed that the proposed Neuro-Fuzzy controller,had better performance on stability,adaptability and real-time capability than the traditional one.【Conclusion】 The designed controller has a strong practical applicability and meets the demands of a smartcar controller.