Selection and Evaluation of Robotic Arm based Conveyor Belts (RACBs) Motions: NARMA(L2)-FO(ANFIS)PD-I based Jaya Optimization Algorithm

Scholars worldwide have shown considerable interest in the industrial sector, mainly due to its abundant resources, which have facilitated the adoption of conveyor belt technologies like Robotic Arm-Based Conveyor Belts (RACBs). RACBs rely on four primary movements: (i.e., joint, motor, gear, and sensor), which can have a significant impact on the overall motions and motion estimation. To optimize these operations, an assistive algorithm has been developed to enhance the effectiveness of motion by achieving favorable gains. However, each motion requires specific criteria for Fractional Order Proportional Integral Derivative (FOPID) controller gains, leading to various challenges. These challenges include the existence of multiple evaluation and selection criteria, the significance of these criteria for each motion, the trade-off between criterion performance for each motion, and determining critical values for the criteria. As a result, the evaluation and selection of the Proposed Jaya optimization algorithm for RACB motion control becomes a complex problem. To address these challenges, this study proposes a novel integrated approach for selecting the Jaya optimization algorithm in different RACB motions. The approach incorporates two evaluation methods: the Nonlinear Autoregressive Moving Average with exogenous inputs (NARMA-L2) controller for Neural Network (NN) weighting of the criteria, and the Adaptive Neuro-Fuzzy Inference System (ANFIS) for selecting the Jaya optimization algorithm. The approach consists of three main phases: RACB-based NARMA-L2 Controller Identification and Pre-processing, Development of NARMA-L2 controller-based NARMA(L2)-FO(ANFIS)PD-I, and Evaluation of FOPID criteria based on JOA. The proposed approach is evaluated based on NARMA(L2)-FO(ANFIS)PD-I that given 0.4074, 0.3156, 0.3724, 0.1898 and 0.2135 for K_p_joint, K_i_motor, K_d_sensor, λ_gear, and µ_N respectively, which verifies the soundness of the proposed methodology.