Razali Muhammad Razmi, Mohd Faudzi Ahmad Athif, Shamsudin Abu Ubaidah, Mohamaddan Shahrol
Faculty of Electrical Engineering, Universiti Teknologi Malaysia, Johor Bahru, Malaysia.
Centre for Artificial Intelligence and Robotics, Universiti Teknologi Malaysia, Kuala Lumpur, Malaysia.
Front Robot AI. 2023 Jan 12;9:1087371. doi: 10.3389/frobt.2022.1087371. eCollection 2022.
Due to the complexity of autonomous mobile robot's requirement and drastic technological changes, the safe and efficient path tracking development is becoming complex and requires intensive knowledge and information, thus the demand for advanced algorithm has rapidly increased. Analyzing unstructured gain data has been a growing interest among researchers, resulting in valuable information in many fields such as path planning and motion control. Among those, motion control is a vital part of a fast, secure operation. Yet, current approaches face problems in managing unstructured gain data and producing accurate local planning due to the lack of formulation in the knowledge on the gain optimization. Therefore, this research aims to design a new gain optimization approach to assist researcher in identifying the value of the gain's product with a qualitative comparative study of the up-to-date controllers. Gains optimization in this context is to classify the near perfect value of the gain's product and processes. For this, a domain controller will be developed based on the attributes of the Fuzzy-PID parameters. The development of the Fuzzy Logic Controller requires information on the PID controller parameters that will be fuzzified and defuzzied based on the resulting 49 fuzzy rules. Furthermore, this fuzzy inference will be optimized for its usability by a genetic algorithm (GA). It is expected that the domain controller will give a positive impact to the path planning position and angular PID controller algorithm that meet the autonomous demand.
由于自主移动机器人需求的复杂性和技术的急剧变化，安全高效的路径跟踪开发变得日益复杂，需要大量的知识和信息，因此对先进算法的需求迅速增加。分析非结构化增益数据一直是研究人员日益关注的问题，这在路径规划和运动控制等许多领域产生了有价值的信息。其中，运动控制是快速、安全运行的关键部分。然而，由于在增益优化知识方面缺乏公式化，当前的方法在管理非结构化增益数据和进行精确的局部规划时面临问题。因此，本研究旨在设计一种新的增益优化方法，通过对最新控制器的定性比较研究，帮助研究人员确定增益乘积的值。在这种情况下，增益优化是对增益乘积和过程的近似完美值进行分类。为此，将基于模糊PID参数的属性开发一个域控制器。模糊逻辑控制器的开发需要PID控制器参数的信息，这些参数将根据生成的49条模糊规则进行模糊化和去模糊化。此外，这种模糊推理将通过遗传算法（GA）对其可用性进行优化。预计该域控制器将对满足自主需求的路径规划位置和角度PID控制器算法产生积极影响。
