Neural Networks Trained via Reinforcement Learning Stabilize Walking of a Three-Dimensional Biped Model With Exoskeleton Applications.
Ontology highlight
ABSTRACT:
SUBMITTER: Liu C
PROVIDER: S-EPMC8378504 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The scale of biological microscopy has increased dramatically over the past ten years, with the development of new modalities supporting collection of high-resolution fluorescence image volumes spanning hundreds of microns if not millimeters. The size and complexity of these volumes is such that quantitative analysis requires automated methods of image processing to identify and characterize individual cells. For many workflows, this process starts with segmentation of nuclei that, due to their ubiquity, ease-of-labeling and relatively simple structure, make them appealing targets for automated detection of individual cells. However, in the context of large, three-dimensional image volumes, nuclei present many challenges to automated segmentation, such that conventional approaches are seldom effective and/or robust. Techniques based upon deep-learning have shown great promise, but enthusiasm for applying these techniques is tempered by the need to generate training data, an arduous task, particularly in three dimensions. Here we present results of a new technique of nuclear segmentation using neural networks trained on synthetic data. Comparisons with results obtained using commonly-used image processing packages demonstrate that DeepSynth provides the superior results associated with deep-learning techniques without the need for manual annotation.
| S-EPMC6892824 | biostudies-literature
Project description:Deep-learning models have become pervasive tools in science and engineering. However, their energy requirements now increasingly limit their scalability1. Deep-learning accelerators2-9 aim to perform deep learning energy-efficiently, usually targeting the inference phase and often by exploiting physical substrates beyond conventional electronics. Approaches so far10-22 have been unable to apply the backpropagation algorithm to train unconventional novel hardware in situ. The advantages of backpropagation have made it the de facto training method for large-scale neural networks, so this deficiency constitutes a major impediment. Here we introduce a hybrid in situ-in silico algorithm, called physics-aware training, that applies backpropagation to train controllable physical systems. Just as deep learning realizes computations with deep neural networks made from layers of mathematical functions, our approach allows us to train deep physical neural networks made from layers of controllable physical systems, even when the physical layers lack any mathematical isomorphism to conventional artificial neural network layers. To demonstrate the universality of our approach, we train diverse physical neural networks based on optics, mechanics and electronics to experimentally perform audio and image classification tasks. Physics-aware training combines the scalability of backpropagation with the automatic mitigation of imperfections and noise achievable with in situ algorithms. Physical neural networks have the potential to perform machine learning faster and more energy-efficiently than conventional electronic processors and, more broadly, can endow physical systems with automatically designed physical functionalities, for example, for robotics23-26, materials27-29 and smart sensors30-32.
| S-EPMC8791835 | biostudies-literature
Project description:Legged robots are better able to adapt to different terrains compared with wheeled robots. However, traditional motion controllers suffer from extremely complex dynamics properties. Reinforcement learning (RL) helps to overcome the complications of dynamics design and calculation. In addition, the high autonomy of the RL controller results in a more robust response to complex environments and terrains compared with traditional controllers. However, RL algorithms are limited by the problems of convergence and training efficiency due to the complexity of the task. Learn and outperform the reference motion (LORM), an RL based framework for gait controlling of biped robot is proposed leveraging the prior knowledge of reference motion. The proposed trained agent outperformed the reference motion and existing motion-based methods. The RL environment was finely crafted for optimal performance, including the pruning of state space and action space, reward shaping, and design of episode criterion. Several improvements were implemented to further improve the training efficiency and performance including: random state initialization (RSI), the noise of joint angles, and a novel improvement based on symmetrization of gait. To validate the proposed method, the Darwin-op robot was set as the target platform and two different tasks were designed: (I) Walking as fast as possible and (II) Tracking specific velocity. In task (I), the proposed method resulted in the walking velocity of 0.488 m/s, with a 5.8 times improvement compared with the original traditional reference controller. The directional accuracy improved by 87.3%. The velocity performance achieved 2× compared with the rated max velocity and more than 8× compared with other recent works. To our knowledge, our work achieved the best velocity performance on the platform Darwin-op. In task (II), the proposed method achieved a tracking accuracy of over 95%. Different environments are introduced including plains, slopes, uneven terrains, and walking with external force, where the robot was expected to maintain walking stability with ideal speed and little direction deviation, to validate the performance and robustness of the proposed method.
| S-EPMC9044250 | biostudies-literature
Project description:Few studies have systematically investigated robust controllers for lower limb rehabilitation exoskeletons (LLREs) that can safely and effectively assist users with a variety of neuromuscular disorders to walk with full autonomy. One of the key challenges for developing such a robust controller is to handle different degrees of uncertain human-exoskeleton interaction forces from the patients. Consequently, conventional walking controllers either are patient-condition specific or involve tuning of many control parameters, which could behave unreliably and even fail to maintain balance. We present a novel, deep neural network, reinforcement learning-based robust controller for a LLRE based on a decoupled offline human-exoskeleton simulation training with three independent networks, which aims to provide reliable walking assistance against various and uncertain human-exoskeleton interaction forces. The exoskeleton controller is driven by a neural network control policy that acts on a stream of the LLRE's proprioceptive signals, including joint kinematic states, and subsequently predicts real-time position control targets for the actuated joints. To handle uncertain human interaction forces, the control policy is trained intentionally with an integrated human musculoskeletal model and realistic human-exoskeleton interaction forces. Two other neural networks are connected with the control policy network to predict the interaction forces and muscle coordination. To further increase the robustness of the control policy to different human conditions, we employ domain randomization during training that includes not only randomization of exoskeleton dynamics properties but, more importantly, randomization of human muscle strength to simulate the variability of the patient's disability. Through this decoupled deep reinforcement learning framework, the trained controller of LLREs is able to provide reliable walking assistance to patients with different degrees of neuromuscular disorders without any control parameter tuning. A universal, RL-based walking controller is trained and virtually tested on a LLRE system to verify its effectiveness and robustness in assisting users with different disabilities such as passive muscles (quadriplegic), muscle weakness, or hemiplegic conditions without any control parameter tuning. Analysis of the RMSE for joint tracking, CoP-based stability, and gait symmetry shows the effectiveness of the controller. An ablation study also demonstrates the strong robustness of the control policy under large exoskeleton dynamic property ranges and various human-exoskeleton interaction forces. The decoupled network structure allows us to isolate the LLRE control policy network for testing and sim-to-real transfer since it uses only proprioception information of the LLRE (joint sensory state) as the input. Furthermore, the controller is shown to be able to handle different patient conditions without the need for patient-specific control parameter tuning.
| S-EPMC10024861 | biostudies-literature
Project description:The spring-loaded inverted pendulum model is similar to human walking in terms of the center of mass (CoM) trajectory and the ground reaction force. It is thus widely used in humanoid robot motion planning. A method that uses a velocity feedback controller to adjust the landing point of a robot leg is inaccurate in the presence of disturbances and a nonlinear optimization method with multiple variables is complicated and thus unsuitable for real-time control. In this paper, to achieve real-time optimization, a CoM-velocity feedback controller is used to calculate the virtual landing point. We construct a touchdown return map based on a virtual landing point and use nonlinear least squares to optimize spring stiffness. For robot whole-body control, hierarchical quadratic programming optimization is used to achieve strict task priority. The dynamic equation is given the highest priority and inverse dynamics are directly used to solve it, reducing the number of optimizations. Simulation and experimental results show that a force-controlled biped robot with the proposed method can stably walk on unknown uneven ground with a maximum obstacle height of 5 cm. The robot can recover from a 5 Nm disturbance during walking without falling.
| S-EPMC7957877 | biostudies-literature
Project description:The purpose of this work is to investigate the accurate trajectory tracking control of a wheeled mobile robot (WMR) based on the slip model prediction. Generally, a nonholonomic WMR may increase the slippage risk, when traveling on outdoor unstructured terrain (such as longitudinal and lateral slippage of wheels). In order to control a WMR stably and accurately under the effect of slippage, an unscented Kalman filter and neural networks (NNs) are applied to estimate the slip model in real time. This method exploits the model approximating capabilities of nonlinear state-space NN, and the unscented Kalman filter is used to train NN's weights online. The slip parameters can be estimated and used to predict the time series of deviation velocity, which can be used to compensate control inputs of a WMR. The results of numerical simulation show that the desired trajectory tracking control can be performed by predicting the nonlinear slip model.
| S-EPMC4965072 | biostudies-literature
Project description:We propose a method of bottom-up coarse-graining, in which interactions within a coarse-grained model are determined by an artificial neural network trained on structural data obtained from multiple atomistic simulations. The method uses ideas of the inverse Monte Carlo approach, relating changes in the neural network weights with changes in average structural properties, such as radial distribution functions. As a proof of concept, we demonstrate the method on a system interacting by a Lennard-Jones potential modeled by a simple linear network and a single-site coarse-grained model of methanol-water solutions. In the latter case, we implement a nonlinear neural network with intermediate layers trained by atomistic simulations carried out at different methanol concentrations. We show that such a network acts as a transferable potential at the coarse-grained resolution for a wide range of methanol concentrations, including those not included in the training set.
| S-EPMC10569054 | biostudies-literature
Project description:The cerebral cortex predicts visual motion to adapt human behavior to surrounding objects moving in real time. Although the underlying mechanisms are still unknown, predictive coding is one of the leading theories. Predictive coding assumes that the brain's internal models (which are acquired through learning) predict the visual world at all times and that errors between the prediction and the actual sensory input further refine the internal models. In the past year, deep neural networks based on predictive coding were reported for a video prediction machine called PredNet. If the theory substantially reproduces the visual information processing of the cerebral cortex, then PredNet can be expected to represent the human visual perception of motion. In this study, PredNet was trained with natural scene videos of the self-motion of the viewer, and the motion prediction ability of the obtained computer model was verified using unlearned videos. We found that the computer model accurately predicted the magnitude and direction of motion of a rotating propeller in unlearned videos. Surprisingly, it also represented the rotational motion for illusion images that were not moving physically, much like human visual perception. While the trained network accurately reproduced the direction of illusory rotation, it did not detect motion components in negative control pictures wherein people do not perceive illusory motion. This research supports the exciting idea that the mechanism assumed by the predictive coding theory is one of basis of motion illusion generation. Using sensory illusions as indicators of human perception, deep neural networks are expected to contribute significantly to the development of brain research.
| S-EPMC5863044 | biostudies-literature
Project description:Human walking is a dynamic, partly self-stabilizing process relying on the interaction of the biomechanical design with its neuronal control. The coordination of this process is a very difficult problem, and it has been suggested that it involves a hierarchy of levels, where the lower ones, e.g., interactions between muscles and the spinal cord, are largely autonomous, and where higher level control (e.g., cortical) arises only pointwise, as needed. This requires an architecture of several nested, sensori-motor loops where the walking process provides feedback signals to the walker's sensory systems, which can be used to coordinate its movements. To complicate the situation, at a maximal walking speed of more than four leg-lengths per second, the cycle period available to coordinate all these loops is rather short. In this study we present a planar biped robot, which uses the design principle of nested loops to combine the self-stabilizing properties of its biomechanical design with several levels of neuronal control. Specifically, we show how to adapt control by including online learning mechanisms based on simulated synaptic plasticity. This robot can walk with a high speed (>3.0 leg length/s), self-adapting to minor disturbances, and reacting in a robust way to abruptly induced gait changes. At the same time, it can learn walking on different terrains, requiring only few learning experiences. This study shows that the tight coupling of physical with neuronal control, guided by sensory feedback from the walking pattern itself, combined with synaptic learning may be a way forward to better understand and solve coordination problems in other complex motor tasks.
| S-EPMC1914373 | biostudies-literature
Project description:The focus of research in biped locomotion has moved toward real-life scenario applications, like walking on uneven terrain, passing through doors, climbing stairs and ladders. As a result, we are witnessing significant advances in the locomotion of biped robots, enabling them to move in hazardous environments while simultaneously accomplishing complex manipulation tasks. Yet, considering walking in an unknown environment, the efficiency of humanoid robots is still far from being comparable with the human. Currently, bipeds are very sensitive to external changes and they have severe constraints for adaptation of walk to conditions from such a complex environment. Promising approaches for efficient generation and realization of walking in a complex environment are based on biological solutions that have been developed for many years of evolution. This work presents one such human-inspired methodology for path planning and realization of biped walk appropriate for motion in a complex unfamiliar environment. Path planning results in calculating clothoid curves that represent well the human-like walking path. The robot walk is realized by the composition of parametric motion primitives. Such an approach enables on-line modification of planned path and walk parameters at any moment, instantly. To establish the relationship between high-level path planner and the low-level joint motion realization, we had to find a way to extract the parameters of the clothoid paths that can be linked with the parameters of the walk and consequently to motion primitive parameters. This enabled the robot to adopt its walking for avoiding the obstacles and for a smooth transition between different paths. In this paper we provide a complete framework that integrates the following components: (i) bio-inspired online path planning, (ii) path-dependent automatic calculation of high-level gait parameters (step length, walking speed, direction, and the height of the foot sole), and (iii) automatic calculation of low-level joint movements and corresponding control terms (driving motor voltage) through the adaptation of motion primitives which realize walking pattern and preserves the dynamic balance of the robot.
| S-EPMC6558152 | biostudies-literature