Obstacle Avoidance and Environmental Adaptability Analysis of Snake-like Robot Based on Deep Learning

Aiming at the problems of high complexity and low accuracy of visual depth map feature recognition, a graph recognition algorithm based on principal component direction depth gradient histogram (pca-hodg) is designed in this study. In order to obtain high-quality depth map, it is necessary to calculate the parallax of the visual image. At the same time, in order to obtain the quantized regional shape histogram, it is necessary to carry out edge detection and gradient calculation on the depth map, then reduce the dimension of the depth map combined with the principal component, and use the sliding window detection method to reduce the dimension again to realize the feature extraction of the depth map. The results show that compared with other algorithms, the pca-hodg algorithm designed in this study improves the average classification accuracy and significantly reduces the average running time. This shows that the algorithm can reduce the running time by reducing the dimension, extract the depth map features more accurately, and has good robustness.

Particle localization using local gradients and its application to nanometer stabilization of a microscope

Accurate localization of single particles plays an increasingly important role in a range of biological techniques, including single molecule tracking and localization-based superresolution microscopy. Such techniques require fast and accurate particle localization algorithms as well as nanometer-scale stability of the microscope. Here, we present a universal method for three-dimensional localization of single labeled and unlabeled particles based on local gradient calculation of microscopy images. The method outperforms current techniques in high noise conditions, and it is capable of nanometer accuracy localization of nano- and micro-particles with sub-ms calculation time. By localizing a fixed particle as fiducial mark and running a feedback loop, we demonstrate its applicability for active drift correction in sensitive nanomechanical measurements such as optical trapping and superresolution imaging. A multiplatform open software package comprising a set of tools for local gradient calculation in brightfield and fluorescence microscopy is shared to the scientific community.

Adjoint Numerical Method for a Multiphysical Inverse Problem of Two-Phase Well Testing in Petroleum Reservoirs

Inverse problem solution is an integral part of data interpretation for well testing in petroleum reservoirs. In case of two-phase well tests with water injection, forward problem is based on the multiphase flow model in porous media and solved numerically. The inverse problem is based on a misfit or likelihood objective function. Adjoint methods have proved robust and efficient for gradient calculation of the objective function in this type of problems. However, if time-lapse electrical resistivity measurements during the well test are included in the objective function, both the forward and inverse problems become multiphysical, and straightforward application of the adjoint method is problematic. In this paper we present a novel adjoint algorithm for the inverse problems considered. It takes into account the structure of cross dependencies between flow and electrical equations and variables, as well as specifics of the equations (mixed parabolic-hyperbolic for flow and elliptic for electricity), numerical discretizations and grids, and measurements in the inverse problem. Decomposition is proposed for the adjoint problem which makes possible step-wise solution of the electric adjoint equations, like in the forward problem, after which a cross-term is computed and added to the right-hand side of the flow adjoint equations at this timestep. The overall procedure provides accurate gradient calculation for the multiphysical objective function while preserving robustness and efficiency of the adjoint methods. Example cases of the adjoint gradient calculation are presented and compared to straightforward difference-based gradient calculation in terms of accuracy and efficiency.

Flexible Transmitter Network

Current neural networks are mostly built on the MP model, which usually formulates the neuron as executing an activation function on the real-valued weighted aggregation of signals received from other neurons. This letter proposes the flexible transmitter (FT) model, a novel bio-plausible neuron model with flexible synaptic plasticity. The FT model employs a pair of parameters to model the neurotransmitters between neurons and puts up a neuron-exclusive variable to record the regulated neurotrophin density. Thus, the FT model can be formulated as a two-variable, two-valued function, taking the commonly used MP neuron model as its particular case. This modeling manner makes the FT model biologically more realistic and capable of handling complicated data, even spatiotemporal data. To exhibit its power and potential, we present the flexible transmitter network (FTNet), which is built on the most common fully connected feedforward architecture taking the FT model as the basic building block. FTNet allows gradient calculation and can be implemented by an improved backpropagation algorithm in the complex-valued domain. Experiments on a broad range of tasks show that FTNet has power and potential in processing spatiotemporal data. This study provides an alternative basic building block in neural networks and exhibits the feasibility of developing artificial neural networks with neuronal plasticity.

A New Hyper-Parameter Optimization Method for Power Load Forecast Based on Recurrent Neural Networks

The selection of the hyper-parameters plays a critical role in the task of prediction based on the recurrent neural networks (RNN). Traditionally, the hyper-parameters of the machine learning models are selected by simulations as well as human experiences. In recent years, multiple algorithms based on Bayesian optimization (BO) are developed to determine the optimal values of the hyper-parameters. In most of these methods, gradients are required to be calculated. In this work, the particle swarm optimization (PSO) is used under the BO framework to develop a new method for hyper-parameter optimization. The proposed algorithm (BO-PSO) is free of gradient calculation and the particles can be optimized in parallel naturally. So the computational complexity can be effectively reduced which means better hyper-parameters can be obtained under the same amount of calculation. Experiments are done on real world power load data,where the proposed method outperforms the existing state-of-the-art algorithms,BO with limit-BFGS-bound (BO-L-BFGS-B) and BO with truncated-newton (BO-TNC),in terms of the prediction accuracy. The errors of the prediction result in different models show that BO-PSO is an effective hyper-parameter optimization method.

An adjoint method for determining the sensitivity of island size to magnetic field variations

An adjoint method to calculate the gradient of island width in stellarators is presented and applied to a set of magnetic field configurations. The underlying method for calculation of the island width is that of Cary & Hanson (Phys. Fluids B, vol. 3, issue 4, 1991, pp. 1006–1014) (with a minor modification), and requires that the residue of the island centre be small. Therefore, the gradient of the residue is calculated in addition. Both the island width and the gradient calculations are verified using an analytical magnetic field configuration introduced by Reiman & Greenside (Comput. Phys. Commun., vol. 43, issue 1, 1986, pp. 157–167). The method is also applied to the calculation of the shape gradient of the width of a magnetic island in a National Compact Stellarator Experiment (NCSX) vacuum configuration with respect to positions on a coil. A gradient-based optimization is applied to a magnetic field configuration studied by Hanson & Cary (Phys. Fluids, vol. 27, issue 4, 1984, pp. 767–769) to minimize stochasticity by adding perturbations to a pair of helical coils. Although only vacuum magnetic fields and an analytical magnetic field model are considered in this work, the adjoint calculation of the island width gradient could also be applied to a magnetohydrodynamic (MHD) equilibrium if the derivative of the magnetic field, with respect to the equilibrium parameters, is known. Using the island width gradient calculation presented here, more general gradient-based optimization methods can be applied to design stellarators with small magnetic islands. Moreover, the sensitivity of the island size may itself be optimized to ensure that coil tolerances, with respect to island size, are kept as high as possible.

Active Machine Learning for Chemical Dynamics Simulations. I. Estimating the Energy Gradient

Ab initio molecular dymamics (AIMD) simulation studies are a direct way to visualize chemical reactions and help elucidate non-statistical dynamics that does not follow the intrinsic reaction coordinate. However, due to the enormous amount of the ab initio energy gradient calculations needed for AIMD, it has been largely restrained to limited sampling and low level of theory (i.e., density functional theory with small basis sets). To overcome this issue, a number of machine learning (ML) methods have been employed to predict the energy gradient of the system of interest. In this manuscript, we outline the theoretical foundations of a novel ML method which trains from a varying set of atomic positions and their energy gradients, called interpolating moving ridge regression (IMRR), and directly predicts the energy gradient of a new set of atomic positions. Several key theoretical findings are presented regarding the inputs used to train IMRR and the predicted energy gradient. A hyperparameter used to guide IMRR is rigorously examined as well. The method is then applied to three bimolecular reactions studied with AIMD, including HBr+ + CO2, H2S + CH, and C4H2 + CH, to demonstrate IMRR’s performance on different chemical systems of different sizes. This manuscript also compares the computational cost of the energy gradient calculation with IMRR vs. ab initio, and the results highlight IMRR as a viable option to greatly increase the efficiency of AIMD.