Search

Proportional-Integral-Derivative (PID) is a simple and intuitive feedback-based control mechanism being useful to track set points and to reject disturbances. A key question in gradient-free optimization is to ascertain whether the class of optimization algorithms based on the difference of vectors generalize reasonably well to tackle a large class of PID control problems. For generalization and practical purposes, it would be desirable to render algorithms being able to tune PID controllers over a diverse and large set of control problems/tasks with minimal human intervention (self-adaptation features), and under tight computational budgets. In this paper, aiming to fill the above-mentioned gap, we propose and investigate the effectiveness of a new class of algorithm based on the difference of vectors and self-adaptation mechanisms for PID tuning. As such, we introduce a new class of Differential Evolution with success-based Particle Adaptations (DEPA), which unifies the principles of difference of vectors, particle schemes and trial/parameter adaptation through archive (memory) mechanisms. Our computational simulations using a large/relevant set of 25 control problem instances (tracking of linear, nonlinear, continuous, and discontinuous trajectories in motor position control, motor velocity control, magnetic levitation, inverted pendulum, crane stabilization), and the comparisons with a large set of closely related optimization algorithms, and their extended adaptive variants (23 optimization algorithms in total) has shown the outperforming benefits of the proposed approach in convergence performance under tight function evaluation budgets (1000 function evaluations). Also, the experiments on a real-world inverted pendulum device show the potential for transferability of the learned gains to unseen situations during training. Furthermore, we evaluated the algorithmic extension and the generalization towards diverse fitness landscapes in the CEC 2017 benchmark suite, showing the attractive/outperforming performance overall problem instances. In particular, the proposed framework performed better in 558 control instances when using ISE as a performance metric, and in 358 control instances when using IAE, ITAE and ITSE as performance metrics. Also, the algorithmic extension for general optimization landscapes performed better than the related algorithms in 182, 215 and 235 problem instances of CEC 2017 benchmark suite for 10, 30 and 50 dimensions, respectively. Our obtained results have the potential to further advance towards developing efficient and self-adaptive optimization algorithms based on the difference of vectors, which may find use in a wider set of optimization and control problems.

A new interconnected translational manipulator is proposed. It is the only interconnected manipulator that makes such motion using revolute joints and three rotary actuators. Rotary joints and actuators are favored practically than their linear counterparts due to their lower price, lower size of installation and higher reliability. The configuration of the proposed manipulator allows it to maintain, to a large extent, the combined merits of serial and parallel manipulators. In contrast to all other existing interconnected manipulators, the proposed manipulator has free-internal-singularity workspace. Using a practical proposed methodology, a balancing system is developed that reduces dramatically the power consumption and facilitates using small-sized-motors. The mobility analysis is carried out using a newly developed methodology suitable for interconnected manipulators. Closed forms for position and velocity kinematics as well as for maximum cuboid workspace are derived. The developed mechanical design is validated by finite element analysis. The controller performance is tested using ADAMS®& MATLAB®/Simulink co-simulation. The results indicate the feasibility of the proposed manipulator and its advantages over existing translational manipulators from engineering as well as economic viewpoints.

Accurately detecting early developmental stages of insect pests (larvae) from off-the-shelf stereo camera sensor data using deep learning holds several benefits for farmers, from simple robot configuration to early neutralization of this less agile but more disastrous stage. Machine vision technology has advanced from bulk spraying to precise dosage to directly rubbing on the infected crops. However, these solutions primarily focus on adult pests and post-infestation stages. This study suggested using a front-pointing red-green-blue (RGB) stereo camera mounted on a robot to identify pest larvae using deep learning. The camera feeds data into our deep-learning algorithms experimented on eight ImageNet pre-trained models. The combination of the insect classifier and the detector replicates the peripheral and foveal line-of-sight vision on our custom pest larvae dataset, respectively. This enables a trade-off between the robot’s smooth operation and localization precision in the pest captured, as it first appeared in the farsighted section. Consequently, the nearsighted part utilizes our faster region-based convolutional neural network-based pest detector to localize precisely. Simulating the employed robot dynamics using CoppeliaSim and MATLAB/SIMULINK with the deep-learning toolbox demonstrated the excellent feasibility of the proposed system. Our deep-learning classifier and detector exhibited 99% and 0.84 accuracy and a mean average precision, respectively.

Mobile robots have recently attracted the attention and applicability in field areas ubiquitously. Within the context of autonomous navigation, path planning is relevant for comfortability, safety, execution time and energy savings. In this paper, we propose an approach to suggest smooth paths from observed robot trajectories by optimizing fitting and smoothness criteria using Differential Evolution with distinct modes of initialization, selection pressure, exploration and exploitation. Our rigorous computational experiments using a relevant set of real-world robot trajectories from the Boe-Bot mobile robot architecture show the feasibility and efficiency of our approach in computing smooth curves, suggesting the superior performance of the greedy initialization scheme based on the triangular convex hull of the robot trajectory, and Differential Evolution with exploitative and parameter adaptation schemes such as Rank-Based Differential Evolution (RBDE), Adaptive Differential Evolution with External Archive (JADE) and Strategy Adaptation Differential Evolution (SADE). Our obtained results offer the building blocks to further advance towards developing data-driven curve fitting and path planning algorithms, which may find use in several real-world applications in Robotics and Operations Research.

Being trajectory tracking key for safe mobile robot navigation, Fuzzy Logic (FL) has been useful in tackling uncertainty and imprecision to realize robust and smooth trajectory tracking. In this paper, we present the Z-number based Fuzzy Logic control for trajectory tracking of differential wheeled mobile robots. The unique point of our approach lies in the ability to encode constraint and reliability in multi-input and multi-output rules, whose antecedent universe considers only the instantaneous measurements of distance and the orientation gaps, and whose consequent universe is computed by the interpolative reasoning and the graded mean integration approach. As a consequence, not only our approach avoids the complexity of encoding error gradients, but also is advantageous to model versatile control rules able to cope with missing observations and noisy inducements on actuators. Our experiments using both physics-based simulations and real-world tests based on a Pioneer 3DX robot architecture have elucidated the superior efficacy and the feasibility of the proposed controller regarding accuracy, robustness, and smoothness compared to other well-known related frameworks such as Fuzzy Logic Type 1, Fuzzy Logic Type 2 and Fuzzy Logic with PID. Our results provide unique insights to realize generalizable algorithms aided by FL and Z-number towards robust trajectory tracking.

Abstract Breast cancer diagnosis has been mostly accomplished through imaging. These methods have great advantages in being able to detect the presence and location of breast cancer. However, it is difficult to distinguish between a benign and malignant tumor located in a deep position because both tumor types look similar. In this paper, tissue including the tumor from skin was vibrated using a compression cylinder, to analyze the frequency difference for distinguishing tissue type. Before distinguishing between a benign and malignant tumor, it is necessary to validate that the difference between normal tissue and tumor can be distinguished. The objective of the study is to validate the feasibility to emphasize the frequency differences in a 10.0 mm or greater deep tumors during vibration by pushing a cylinder towards the deep tumor. A phantom model and finite element analysis model were constructed to simulate the breast. In the experiment, air was injected into the phantom and the displacement was measured. The frequency response for distinction of tissue types was analyzed and it was found that the displacement difference rate was over 50% at a frequency of 130 Hz when the cylinder was pushed into the sample as opposed to when not pushed in. Changes in displacement were measured according to the distance between the tumor and vibration point using finite element analysis. When the measurement and vibration points were on the center of the tumor, the difference in the resonance point was at its largest (5.5 Hz). Results show that the position of a tumor could be easily and rapidly detected by vibrations from a cylinder pushed into the diagnostic site.

This paper presents a method to inspect the interior of a winding small gas pipe using a hollow guide wire. There is no conventional method to insert an endoscope into an 8-bend 25-mm-diameter gas pipe within 2 hours. In medical practice, a guide wire inserted in advance enables insertion of a catheter into a vessel. However, it is impossible to insert a normal guide wire into a gas pipe because the wire buckles in the pipe. Thus, we designed a hollow guide wire with a small front diameter and large rear diameter, making the front soft and the rear stiff. This guide wire can be inserted without buckling or meandering. First, we measured mechanical properties such as the torsional spring constant and damping coefficient of the wire and the frictional coefficient between the pipe and wire. Second, we conducted an experiment inserting guide wires with various tip pitches, front lengths, front outer diameters, and rear outer diameters. Third, we analyzed the insertion distance by simulating guide wire insertion using the Lagrange method, and optimized the guide wire shape via the response surface method. Finally, the optimized guide wires were tested experimentally to validate the analysis. As a result, an optimized guide wire and an endoscope can both be inserted into a gas pipe and removed within just 3 minutes.