Artificial Potential Research Articles

Artificial neural network potentials for mechanics and fracture dynamics of two-dimensional crystals ** ** Notice: This manuscript has been authored by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a nonexclusive, paid-up, irrevocable, world-wide license

Understanding the mechanics and failure of materials at the nanoscale is critical for their engineering and applications. The accurate atomistic modeling of brittle failure with crack propagation in covalent crystals requires a quantum mechanics-based description of individual bond-breaking events. Artificial neural network potentials (NNPs) have emerged to overcome the traditional, physics-based modeling tradeoff between accuracy and accessible time and length scales. Previous studies have shown successful applications of NNPs for describing the structure and dynamics of molecular systems and amorphous or liquid phases of materials. However, their application to deformation and failure processes in materials is still uncommon. In this study, we discuss the apparent limitations of NNPs for the description of deformation and fracture under loadings and propose a way to generate and select training data for their employment in simulations of deformation and fracture simulations of crystals. We applied the proposed approach to 2D crystalline graphene, utilizing the density-functional tight-binding method for more efficient and extensive data generation in place of density functional theory. Then, we explored how the data selection affects the accuracy of the developed artificial NNPs. It revealed that NNP’s reliability should not only be measured based on the total energy and atomic force comparisons for reference structures but also utilize comparisons for physical properties, e.g. stress–strain curves and geometric deformation. In sharp contrast to popular reactive bond order potentials, our optimized NNP predicts straight crack propagation in graphene along both armchair and zigzag (ZZ) lattice directions, as well as higher fracture toughness of ZZ edge direction. Our study provides significant insight into crack propagation mechanisms on atomic scales and highlights strategies for NNP developments of broader materials.

Particle Swarm Algorithm Path-Planning Method for Mobile Robots Based on Artificial Potential Fields

Path planning is an important part of the navigation control system of mobile robots since it plays a decisive role in whether mobile robots can realize autonomy and intelligence. The particle swarm algorithm can effectively solve the path-planning problem of a mobile robot, but the traditional particle swarm algorithm has the problems of a too-long path, poor global search ability, and local development ability. Moreover, the existence of obstacles makes the actual environment more complex, thus putting forward more stringent requirements on the environmental adaptation ability, path-planning accuracy, and path-planning efficiency of mobile robots. In this study, an artificial potential field-based particle swarm algorithm (apfrPSO) was proposed. First, the method generates robot planning paths by adjusting the inertia weight parameter and ranking the position vector of particles (rPSO), and second, the artificial potential field method is introduced. Through comparative numerical experiments with other state-of-the-art algorithms, the results show that the algorithm proposed was very competitive.

Spacecraft Rendezvous and Docking Using the Explicit Reference Governor Approach

This article investigates the autonomous spacecraft rendezvous and docking problem in the presence of space obstacle, path constraints, and thrust limitation constraint. By introducing the frame of explicit reference governor (ERG), a combination scheme of guidance and control is proposed to guarantee the system stabilization and constraints satisfaction. Specifically, an artificial potential function (APF) method is employed to guide a collision-free trajectory, and then the constraints are satisfied via limiting the states within the safe invariant set. Furthermore, a simplified method is proposed to obtain the maximum bound of the Lyapunov-based invariant sets, which ensures input bounded and collision avoidance. System convergence under the potential field is proved through the Lyapunov stability analysis. Numerical simulation results demonstrate the comprehensive validation and good performance of the proposed method.

Classical and Heuristic Approaches for Mobile Robot Path Planning: A Survey

The most important research area in robotics is navigation algorithms. Robot path planning (RPP) is the process of choosing the best route for a mobile robot to take before it moves. Finding an ideal or nearly ideal path is referred to as “path planning optimization.” Finding the best solution values that satisfy a single or a number of objectives, such as the shortest, smoothest, and safest path, is the goal. The objective of this study is to present an overview of navigation strategies for mobile robots that utilize three classical approaches, namely: the roadmap approach (RM), cell decomposition (CD), and artificial potential fields (APF), in addition to eleven heuristic approaches, including the genetic algorithm (GA), ant colony optimization (ACO), artificial bee colony (ABC), gray wolf optimization (GWO), shuffled frog-leaping algorithm (SFLA), whale optimization algorithm (WOA), bacterial foraging optimization (BFO), firefly (FF) algorithm, cuckoo search (CS), and bat algorithm (BA), which may be used in various environmental situations. Multiple issues, including dynamic goals, static and dynamic environments, multiple robots, real-time simulation, kinematic analysis, and hybrid algorithms, are addressed in a different set of articles presented in this study. A discussion, as well as thorough tables and charts, will be presented at the end of this work to help readers understand what types of strategies for path planning are developed for use in a wide range of ecological contexts. Therefore, this work’s main contribution is that it provides a broad view of robot path planning, which will make it easier for scientists to study the topic in the near future.

A REVIEW OF METHODOLOGIES FOR PATH PLANNING AND OPTIMIZATION OF MOBILE ROBOTS

Abstract: This research paper provides a comprehensive review of methodologies for path planning and optimization of mobile robots. With the rapid development of robotics technology, path planning and optimization have become fundamental areas of research for achieving efficient and safe autonomous robot navigation. In this paper, we review the classic and state-of-the-art techniques of path planning and optimization, including artificial potential fields, A* algorithm, Dijkstra's algorithm, genetic algorithm, swarm intelligence, and machine learning-based methods. We analyze the strengths and weaknesses of each approach and discuss their application scenarios. Moreover, we identify the challenges and open problems in this field, such as dealing with dynamic environments and real-time constraints. This paper serves as a comprehensive reference for researchers and practitioners in the robotics community, providing insights into the latest trends and developments in path planning and optimization for mobile robots.

Revisiting the stable structures of gold clusters: Au n (n = 16–25) by artificial neural network potential

Identifying the stable structures of gold (Au) clusters is a huge challenge in cluster science. In this work, we have searched the ground-state structures of neutral Au n (n = 16–25) clusters using the potential of an artificial neural network (ANN) trained with density functional theory (DFT) data. Compared with the DFT data, the root mean square error of binding energy predicted by the ANN potential is about 8.66 meV/atom. Applying the ANN potential to search the ground-state structures by comprehensive genetic algorithm, we have found several new candidates of Au18, Au22, and Au23, which have not been previously reported. Au18 has a hollow cage structure, whereas Au22 and Au23 are flat cage structures. From the electronic analysis, we elucidate the stability mechanism of the newly found structures that are associated with the electronic shell closure of superatomic orbitals. Additonally, we also clarified how to clean a database to train an efficient ANN potential in detail. Overall, this work proves that applying machine learning to the description of atomic interactions can accelerate the search of ground-state structures of clusters and help to find new candidates for stable cluster structures.

Online Feet Potential Fields for Quadruped Robots Navigation in Harsh Terrains

Quadruped robots have garnered significant attention in recent years due to their ability to navigate through challenging terrains. Among the various environments, agriculture fields are particularly difficult for legged robots, given the variability of soil types and conditions. To address this issue, this study proposes a novel navigation strategy that utilizes ground reaction forces to calculate online artificial potential fields, which are then applied to the robot’s feet to avoid low-traversability regions. The strategy also incorporates the net vector of the attractive potential field towards the goal and the repulsive field to avoid slippery regions, which dynamically adjusts the quadruped’s gait. A realistic simulation environment validates the proposed navigation framework with case studies on randomly generated terrains. A comprehensive comparison with baseline navigation methods is conducted to assess the effectiveness of the proposed approach.

A REVIEW OF METHODOLOGIES FOR PATH PLANNING AND OPTIMIZATION OF MOBILE ROBOTS

This research paper provides a comprehensive review of methodologies for path planning and optimization of mobile robots. With the rapid development of robotics technology, path planning and optimization have become fundamental areas of research for achieving efficient and safe autonomous robot navigation. In this paper, we review the classic and state-of-the-art techniques of path planning and optimization, including artificial potential fields, A* algorithm, Dijkstra's algorithm, genetic algorithm, swarm intelligence, and machine learning-based methods. We analyze the strengths and weaknesses of each approach and discuss their application scenarios. Moreover, we identify the challenges and open problems in this field, such as dealing with dynamic environments and real-time constraints. This paper serves as a comprehensive reference for researchers and practitioners in the robotics community, providing insights into the latest trends and developments in path planning and optimization for mobile robots.

Artificial Action Potential and Ionic Power Device Inspired by Ion Channels and Excitable Cell (Adv. Sci. 16/2023)

Artificial Action Potential and Ionic Power Device Inspired by Ion Channels and Excitable Cell (Adv. Sci. 16/2023)

Specific Tracking Control of Rotating Target Spacecraft Under Safe Motion Constraints

This paper considers a space mission that a chaser spacecraft approaches and tracks a rotating target at close range under safe motion constraints. The control objective is to keep the chaser hovering at the desired relative position above the specific point of the target and ensure the chaser's line-of-sight pointing to the target. During the proximity process of position and attitude maneuver, two kinds of safe geometric constraints are essential to be considered. One is collision avoidance for position constraint, and the other is the pointing forbidden zone for attitude constraint. Superquadrics curved surface is used to describe the position forbidden. Pointing forbidden is represented by inequality constraint based on SO(3). Combining the artificial potential function with backstepping control methodology and adaptive sliding-mode, a safety robust control law is designed. The proposed control scheme guarantees the chaser to track the target under safe geometric constraints. In addition, by using the dynamic specific point under observation, the proposed control scheme can be applied to other observation missions like close flying around. Finally, simulation results illustrate the effectiveness of the control scheme.

Leveraging complexity frameworks to refine theories of engagement: Advancing self‐regulated learning in the age of artificial intelligence

AbstractCapturing evidence for dynamic changes in self‐regulated learning (SRL) behaviours resulting from interventions is challenging for researchers. In the current study, we identified students who were likely to do poorly in a biology course and those who were likely to do well. Then, we randomly assigned a portion of the students predicted to perform poorly to a science of learning to learn intervention where they were taught SRL study strategies. Learning outcome and log data (257 K events) were collected from n = 226 students. We used a complex systems framework to model the differences in SRL including the amount, interrelatedness, density and regularity of engagement captured in digital trace data (ie, logs). Differences were compared between students who were predicted to (1) perform poorly (control, n = 48), (2) perform poorly and received intervention (treatment, n = 95) and (3) perform well (not flagged, n = 83). Results indicated that the regularity of students' engagement was predictive of course grade, and that the intervention group exhibited increased regularity in engagement over the control group immediately after the intervention and maintained that increase over the course of the semester. We discuss the implications of these findings in relation to the future of artificial intelligence and potential uses for monitoring student learning in online environments. Practitioner notesWhat is already known about this topic Self‐regulated learning (SRL) knowledge and skills are strong predictors of postsecondary STEM student success. SRL is a dynamic, temporal process that leads to purposeful student engagement. Methods and metrics for measuring dynamic SRL behaviours in learning contexts are needed. What this paper adds A Markov process for measuring dynamic SRL processes using log data. Evidence that dynamic, interaction‐dominant aspects of SRL predict student achievement. Evidence that SRL processes can be meaningfully impacted through educational intervention. Implications for theory and practice Complexity approaches inform theory and measurement of dynamic SRL processes. Static representations of dynamic SRL processes are promising learning analytics metrics. Engineered features of LMS usage are valuable contributions to AI models.

An artificial neural network potential for uranium metal at low pressures

Based on machine learning, the high-dimensional fitting of potential energy surfaces under the framework of first principles provides density-functional accuracy of atomic interaction potential for high-precision and large-scale simulation of alloy materials. In this paper, we obtained the high-dimensional neural network (NN) potential function of uranium metal by training a large amount of first-principles calculated data. The lattice constants of uranium metal with different crystal structures, the elastic constants, and the anisotropy of lattice expansion of alpha-uranium obtained based on this potential function are highly consistent with first-principles calculation or experimental data. In addition, the calculated formation energy of vacancies in alpha- and beta-uranium also matches the first-principles calculation. The calculated site of the most stable self-interstitial and its formation energy is in good agreement with the findings from density functional theory (DFT) calculations. These results show that our potential function can be used for further large-scale molecular dynamics simulation studies of uranium metal at low pressures, and provides the basis for further construction of potential model suitable for a wide range of pressures.

Enhanced multi-agent systems formation and obstacle avoidance (EMAFOA) control algorithm

Most of the multi-agent formation and obstacle avoidance algorithms in the literature are computationally expensive and/or presents an ad hoc method for a specific type of systems. A general inexpensive, in term of computational complexity, multi-agent formation and obstacle avoidance algorithm is modeled and designed in this paper. The method builds on the leader-follower strategy for the simplicity and applicability for a wide range of the engineering systems. A novel artificial potential function (APF) is proposed. The proposed function has unique attributes which differentiate it superior to what is reported in the literature. The proposed algorithm can be applied for first order systems, second order systems, and non-holonomic systems. Also, the proposed method is free of local minima problem and oscillations. In addition, a novel multi-agent system formation control design is proposed in this work. The proposed design allows to embed any formation in the system paradigm and reduces complexity. Moreover, the stability of the proposed method is investigated in terms of Riccati equation and Lyapunov method. Furthermore, to show the effectiveness of the proposed method it applied and simulated for a second order leader and one follower system with specific formation. Then, a five second order agents with leader in a circular formation avoiding simple and complex obstacles are also introduced with different scenarios. Finally, twenty agents with square formation with two obstacles are simulated and investigated.

UV-A for Tailoring the Nutritional Value and Sensory Properties of Leafy Vegetables

This study aims to expand the artificial lighting potential of controlled environment cultivations systems by introducing UV-A (~315–400 nm) wavelengths into the traditional, visible spectrum lighting, seeking to improve the nutritional and sensory value of cultivated leafy vegetables. The experiment was conducted in a closed climate-controlled chamber, maintaining 21/17 °C day/night temperature, ~55% relative humidity, and a 16 h photo/thermo period. Several genotypes of leafy vegetables, red and green leaf lettuce cultivars, mustard, and kale were cultivated under 250 µmol m−2s−1 basal LED lighting, supplemented by 385 nm UV-A or 405 UV-A/violet wavelengths for 1.1 mW cm−2 for 12 h photoperiod for the whole cultivation cycle. The results show that UV-A/violet light impacts on leafy vegetable growth, free radical scavenging activity, sugar, and phytochemical (α tocopherol, α + β carotenes, epicatechin, rosmarinic and chicoric acid contents) are species-specific, and do not correlate with untrained consumer’s sensory evaluation scores. The 405 nm light is preferable for higher antioxidant and/or sensory properties of kale, mustard, and green leaf lettuces, but both UV-A wavelengths reduce growth parameters in red leaf lettuce.

Emergency Collision Avoidance and Mitigation Using Model Predictive Control and Artificial Potential Function

Although extensive research in emergency collision avoidance has been carried out for straight or curved roads in a highway scenario, a general method that could be implemented for all road environments has not been thoroughly explored. Moreover, most current algorithms don't consider collision mitigation in an emergency. This functionality is essential since the problem may have no feasible solution. We propose a safe controller using model predictive control and artificial potential function to address these problems. A new artificial potential function inspired by line charge is proposed as the cost function for our model predictive controller. The vehicle dynamics and actuator limitations are set as constraints. The new artificial potential function considers the shape of all objects. In particular, the artificial potential function we proposed has the flexibility to fit the shape of the road structures, such as the intersection. We could also realize collision mitigation for a specific part of the vehicle by increasing the charge quantity at the corresponding place. We have tested our methods in 192 cases from 8 different scenarios in simulation with two different models. The simulation results show that the success rate of the proposed safe controller is 20 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> higher than using HJ-reachability with system decomposition by using a unicycle model. It could also decrease 43 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> of collision that happens at the pre-assigned part. The method is further validated in a dynamic bicycle model.

ænet-PyTorch: A GPU-supported implementation for machine learning atomic potentials training.

In this work, we present ænet-PyTorch, a PyTorch-based implementation for training artificial neural network-based machine learning interatomic potentials. Developed as an extension of the atomic energy network (ænet), ænet-PyTorch provides access to all the tools included in ænet for the application and usage of the potentials. The package has been designed as an alternative to the internal training capabilities of ænet, leveraging the power of graphic processing units to facilitate direct training on forces in addition to energies. This leads to a substantial reduction of the training time by one to two orders of magnitude compared to the central processing unit implementation, enabling direct training on forces for systems beyond small molecules. Here, we demonstrate the main features of ænet-PyTorch and show its performance on open databases. Our results show that training on all the force information within a dataset is not necessary, and including between 10% and 20% of the force information is sufficient to achieve optimally accurate interatomic potentials with the least computational resources.

Artificial potential incorporated adaptive fixed-time sliding mode control for target capture

With the increasing development of space missions, techniques for space non-cooperative target capture are evolving into new stage with higher control requirements under various working scenarios. Imitation Human Hand Capture is an effective rigid envelope capture method and could grasp non-cooperative targets without any cooperative feature points. In order to guarantee the safety and increase the success rate of IHHC, an adaptive fixed-time sliding mode frame based on IHHC is proposed, in which the Imitation Human Hand Capture governs the stable grasp and the sliding mode frame governs the control performance. In the proposed control frame, a fixed-time sliding mode control method is employed and the settling time can be adaptively tuned with the artificial potential function formulated according to the base-target distance and the envelope state of the arms. The proposed control frame is verified through Lyapunov function and simulation tests. The results demonstrate the effectiveness of the proposed control frame.

Composite neural learning based appointed-time safe approach control under full-state constraints

This article investigates composite learning based appointed-time control strategy for spacecraft rendezvous and docking subject to model uncertainties, input saturation, and actuator faults. A novel prescribed performance function (PPF) is designed to constraint all system states and ensure the property of controller insensitive to initial conditions. Combining the artificial potential function (APF) method with transformed error states, a six-degree-of-freedom (6-DOF) coupled relative motion model with full-state and motion constraints is proposed. To obtain higher control accuracy and address the problem of complex calculations in network weight update, a low-complexity composite learning based neural network (NN) weight update law is investigated. The appointed-time controller is designed by utilizing backstepping method, and the dynamic surface technique is used to solve the “dimension explosion” problem. The proposed controller guarantees the stability of the system states in appointed time without the information of target. Three simulation scenarios are carried out to validate the effectiveness and robustness of the proposed controller.

Artificial Action Potential and Ionic Power Device Inspired by Ion Channels and Excitable Cell.

In vivo, the membrane potential of the excitable cell working by ion gradients plays a significant role in bioelectricity generation and nervous system operation. Conventional bioinspired power systems generally have adopted ion gradients, but overlook the functions of ion channels and Donnan effect to generate efficient ion flow in the cell. Here, cell-inspired ionic power device implementing the Donnan effect using multi-ions and monovalent ion exchange membranes as artificial ion channels is realized. Different ion-rich electrolytes on either side of the selective membrane generate the ion gradient potentials with high ionic currents and reduce the osmotic imbalance of the membrane. Based on this device, the artificial neuronal signaling is presented by the mechanical switching system of the ion selectivity like mechanosensitive ion channels in a sensory neuron. Compared with reverse electrodialysis, which requires a low concentration, a high-power device with ten times the current and 8.5 times the power density is fabricated. This device activates grown muscle cells by increasing power through serial connection like an electric eel, and shows the possibility of an ion-based artificial nervous system.

The large-time behavior of the Vlasov alignment model with power-law or Riesz potentials

We consider the Vlasov alignment model with pairwise attractive potentials. By using an artificial potential energy representing the velocity alignment, we construct a new Lyapunov functional, and then establish the large time behavior of the Vlasov alignment model with any regular power-law potential. Compared with previous results, which focus on the low order power-law potential case, we give a better estimate of convergence rates. More importantly, the consensus with polynomial rates is also established for the high order power-law potential case. In the meanwhile, for the Riesz and logarithmic potential case, the non-existence of flocking is proved. Besides, the above large time behavior also holds for the hydrodynamic model.