Closed-loop Experiments Research Articles

Research on nonlinear instability of blended-wing-body aircraft based on experimental bifurcation analysis

Focusing on the nonlinear instability problem of the blended-wing-body aircraft, first, numerical bifurcation analysis is adopted to obtain catastrophe points and branches where the elevator is taken as a continuous parameter. Second, based on virtual flight tests and bifurcation theory, an experimental bifurcation analysis method is developed. Open-loop and closed-loop experiments are carried out, and the corresponding experimental bifurcation diagrams are obtained. The comparative analysis shows that the experimental bifurcation analysis enables the aircraft to make a response that is close to reality under the influence of nonlinear and unsteady aerodynamics and has a more intuitive embodiment of instability characteristics, providing an accurate prediction of departure. The open-loop experiment predicts a sudden longitudinal nose-up at an angle of attack of 7.4°. The closed-loop experiment can transform the open-loop unstable branch into a stable branch. However, roll instability still occurs at a critical angle of attack of 16.0°. These research results provide important dynamic information for the stable flight and the design of the controller.

Experimental demonstration of the production of poly(oxymethylene) dimethyl ethers from methanolic formaldehyde solutions in a closed-loop mini-plant

Poly(oxymethylene) dimethyl ethers of chain length 3-5 ( ▪ ) are discussed as synthetic diesel fuels due to their potential to significantly reduce soot emissions while presenting physicochemical properties similar to conventional diesel fuels. A recently developed process to produce ▪ directly from methanol and aqueous formaldehyde is a promising way to avoid expensive intermediates such as trioxane, methylal, or dimethyl ether. The process consists of a reactor, a distillation sequence, and a membrane unit. The first distillation column is particularly challenging due to the reactive character of the separation, the high number of components present, and the limited solubility of formaldehyde. Up to now, the feasibility of this separation has yet to be demonstrated. This work presents closed-loop experiments in a demonstration plant erected at the Campus Straubing of the Technical University of Munich with closing recycle. In the distillation step, ▪ with with very small impurities of around 800 ppm of formaldehyde was obtained. The experimental results were compared with simulations based on a reactive equilibrium stage model. The simulations of the temperature and composition profiles for the majority of the components are in line with the experiments. Although the membrane separation exhibited weaker water selectivity than in previous tests, it was sufficient to overcome distillation boundaries. Additionally, this study assesses the potential for solid precipitation, explores trade-offs in product quality, and discusses overall mass balances of the process, demonstrating the overall feasibility of the process.

Gate-set evaluation metrics for closed-loop optimal control on nitrogen-vacancy center ensembles in diamond

A recurring challenge in quantum science and technology is the precise control of their underlying dynamics that lead to the desired quantum operations, often described by a set of quantum gates. These gates can be subject to application-specific errors, leading to a dependence of their controls on the chosen circuit, the quality measure and the gate-set itself. A natural solution would be to apply quantum optimal control in an application-oriented fashion. In turn, this requires the definition of a meaningful measure of the contextual gate-set performance. Therefore, we explore and compare the applicability of quantum process tomography, linear inversion gate-set tomography, randomized linear gate-set tomography, and randomized benchmarking as measures for closed-loop quantum optimal control experiments, using a macroscopic ensemble of nitrogen-vacancy centers in diamond as a test-bed. Our work demonstrates the relative trade-offs between those measures and how to significantly enhance the gate-set performance, leading to an improvement across all investigated methods.

NeuroART: Real-Time Analysis and Targeting of Neuronal Population Activity during Calcium Imaging for Informed Closed-Loop Experiments.

Two-photon calcium imaging allows for the activity readout of large populations of neurons at single cell resolution in living organisms, yielding new insights into how the brain processes information. Holographic optogenetics allows us to trigger activity of this population directly, raising the possibility of injecting information into a living brain. Optogenetic triggering of activity that mimics "natural" information, however, requires identification of stimulation targets based on real-time analysis of the functional network. We have developed NeuroART (Neuronal Analysis in Real Time), software that provides real-time readout of neuronal activity integrated with downstream analysis of correlations and synchrony and of sensory metadata. On the example of auditory stimuli, we demonstrate real-time inference of the contribution of each neuron in the field of view to sensory information processing. To avoid the limitations of microscope hardware and enable collaboration of multiple research groups, NeuroART taps into microscope data streams without the need for modification of microscope control software and is compatible with a wide range of microscope platforms. NeuroART also integrates the capability to drive a spatial light modulator (SLM) for holographic photostimulation of optimal stimulation targets, enabling real-time modification of functional networks. Neurons used for photostimulation experiments were extracted from Sprague Dawley rat embryos of both sexes.

Penetration State Recognition during Laser Welding Process Control Based on Two-Stage Temporal Convolutional Networks.

Vision-based laser penetration control has become an important research area in the field of welding quality control. Due to the complexity and large number of parameters in the monitoring model, control of the welding process based on deep learning and the reliance on long-term information for penetration identification are challenges. In this study, a penetration recognition method based on a two-stage temporal convolutional network is proposed to realize the online process control of laser welding. In this paper, a coaxial vision welding monitoring system is built. A lightweight segmentation model, based on channel pruning, is proposed to extract the key features of the molten pool and the keyhole from the clear molten pool keyhole image. Using these molten pool and keyhole features, a temporal convolutional network based on attention mechanism is established. The recognition method can effectively predict the laser welding penetration state, which depends on long-term information. In addition, the penetration identification experiment and closed-loop control experiment of unequal thickness plates are designed. The proposed method in this study has an accuracy of 98.96% and an average inference speed of 20.4 ms. The experimental results demonstrate that the proposed method exhibits significant performance in recognizing the penetration state from long sequences of welding image signals, adjusting welding power, and stabilizing welding quality.

A novel design of a MEMS resonant accelerometer with adjustable sensitivity

This paper presents the design and experimental evaluation of a silicon micro-machined resonant accelerometer featuring adjustable sensitivity. By integrating an electrostatic tuning module into the fundamental accelerometer structure, dynamic sensitivity adjustment becomes feasible, leveraging the softening effect of electrostatic negative stiffness to optimize range, noise, and bandwidth. Notably, the electrostatic tuning module integrates seamlessly with the core accelerometer structure, minimizing structural alterations. Through theoretical analysis and finite element simulation of the electrostatic negative stiffness principle, we have designed a novel accelerometer with adjustable sensitivity, which can enhance the sensitivity and reduces the bias-instability of the accelerometer with a relatively small adjustment voltage, without increasing structural complexity. The performance of the accelerometer was assessed through open-loop, closed-loop, and dynamic experiments, revealing that sensitivity increased from 843 Hz/g to 2611 Hz/g within a linear range of ±1 g when employing a sensitivity-enhancing bias voltage of 9 V. Moreover, the bias-instability is lowered down from 17.3 μg to 6.8 μg. This design offers a promising avenue for sensitivity tuning in MEMS resonant accelerometers.

Closed-loop transfer enables artificial intelligence to yield chemical knowledge.

Artificial intelligence-guided closed-loop experimentation has emerged as a promising method for optimization of objective functions1,2, but the substantial potential of this traditionally black-box approach to uncovering new chemical knowledge has remained largely untapped. Here we report the integration of closed-loop experiments with physics-based feature selection and supervised learning, denoted as closed-loop transfer (CLT), to yield chemical insights in parallel with optimization of objective functions. CLT was used to examine the factors dictating the photostability in solution of light-harvesting donor-acceptor molecules used in a variety of organic electronics applications, and showed fundamental insights including the importance of high-energy regions of the triplet state manifold. This was possible following automated modular synthesis and experimental characterization of only around 1.5% of the theoretical chemical space. This physics-informed model for photostability was strengthened using multiple experimental test sets and validated by tuning the triplet excited-state energy of the solvent to break out of the observed plateau in the closed-loop photostability optimization process. Further applications of CLT to additional materials systems support the generalizability of this strategy for augmenting closed-loop strategies. Broadly, these findings show that combining interpretable supervised learning models and physics-based features with closed-loop discovery processes can rapidly provide fundamental chemical insights.

Dual-reference adaptive feedforward control of a magnetostrictive device for synchronous micropositioning and microvibration isolation

In this paper, we propose a novel dual-reference adaptive feedforward controller to realize synchronous micropositioning and microvibration isolation on a magnetostrictive device. The scheme of the proposed adaptive feedforward controller and its differences from the traditional single-reference controller are briefly introduced. The desired trajectory and available external disturbance as two input references are utilized to develop the proposed controller. The dynamics compensator is constructed based on the modified filtered-x normalized least mean square (MFxNLMS) algorithm with the discrete cosine transform (DCT) technique. The asymmetric hysteresis compensator is modeled via the arctangent-polynomial modified Prandtl–Ishlinskii (APMPI) model. The experimental setup is built, and the closed-loop control experiment is carried out according to the designed experimental process. Comparison of experimental results show that the proposed dual-reference DCT-MFxNLMS controller behaves better than the single-reference DCT-MFxNLMS and proportional-derivative-derivative (PID) integrated controller for the synchronous micropositioning and vibration isolation cases. Moreover, by the dual-reference DCT-MFxNLMS controller, the vibration isolation ratio enhances and the tracking bandwidth increases within the interest of frequency bandwidth, compared with those of open-loop system, respectively.

Coupled hysteresis and resonance control of three-degree-of-freedom tip-tilt-piston piezoelectric stage

In order to solve the coupled hysteresis and resonance problems in three-degree-of-freedom tip-tilt-piston piezoelectric stage, a coupled hysteresis model and a coupled integral resonance compensation model are designed. The inverse coupled hysteresis feedforward control combined multi-axis coupled resonance control and high-gain integral tracking controller are used to improve the tracking speed and positioning accuracy of the stage. Firstly, a kinematic model of the stage is built to translate the three-degree-of-freedom motion of the end-effector into the outputs of three piezoelectric actuators. Then, a coupled hysteresis model based on the rate-dependent Prandtl–Ishlinskii model is established and parameterized to simultaneously compensate for the hysteresis of a single piezoelectric actuator and the coupled hysteresis between multiple piezoelectric actuators. The effectiveness of the model is verified by open-loop and closed-loop inverse model feedforward control experiments. Afterwards, a coupled integral resonance compensation model is also established and parameterized to suppresses the resonance and expand the system bandwidth. Finally, the proposed controller is used to conduct tracking experiments on a composite signal, and the results show that the accuracy and speed of trajectory tracking are further improved, and the mechanical coupling between axes are significantly reduced.

Bridging model and experiment in systems neuroscience with Cleo: the Closed-Loop, Electrophysiology, and Optophysiology simulation testbed.

Systems neuroscience has experienced an explosion of new tools for reading and writing neural activity, enabling exciting new experiments such as all-optical or closed-loop control that effect powerful causal interventions. At the same time, improved computational models are capable of reproducing behavior and neural activity with increasing fidelity. Unfortunately, these advances have drastically increased the complexity of integrating different lines of research, resulting in the missed opportunities and untapped potential of suboptimal experiments. Experiment simulation can help bridge this gap, allowing model and experiment to better inform each other by providing a low-cost testbed for experiment design, model validation, and methods engineering. Specifically, this can be achieved by incorporating the simulation of the experimental interface into our models, but no existing tool integrates optogenetics, two-photon calcium imaging, electrode recording, and flexible closed-loop processing with neural population simulations. To address this need, we have developed Cleo: the Closed-Loop, Electrophysiology, and Optophysiology experiment simulation testbed. Cleo is a Python package enabling injection of recording and stimulation devices as well as closed-loop control with realistic latency into a Brian spiking neural network model. It is the only publicly available tool currently supporting two-photon and multi-opsin/wavelength optogenetics. To facilitate adoption and extension by the community, Cleo is open-source, modular, tested, and documented, and can export results to various data formats. Here we describe the design and features of Cleo, validate output of individual components and integrated experiments, and demonstrate its utility for advancing optogenetic techniques in prospective experiments using previously published systems neuroscience models.

Development of miniaturized J-T cooler with thin-plate type sorption compressor for 5 K cooling

A sorption J-T cooler for a cooling temperature of 5 K can be useful to cool the sensitive detectors and calorimeters, due to its vibration-free characteristic. It is a J-T cooler driven by the sorption compressor, that utilizes the adsorption phenomenon to create the pressure gradient. To maximize the mass flow rate, the switchless thin-plate type cell is adopted for the sorption compressor. The coiled tube-in-tube heat exchanger is fabricated to minimize the overall size of the cooler. After precooling the experimental apparatus with a two-stage G-M cooler, the open-loop test is performed to assess the mass flow rate characteristics of the J-T restrictor and the background heat ingress. In the closed-loop experiments, the cooling temperature below 5 K is achieved by the sorption compressor without heat load. The nominal mass flow rate from the sorption compressor is 1.3 mg/s with the pressure ratio between 5.9 and 6.6. Subsequently, the model of the heat exchanger is utilized to determine the maximum cooling capacity according to the inlet temperature of the high-pressure stream. Furthermore, the effectiveness and the overall COP of the sorption J-T cooler are analyzed. The maximum cooling capacity at 5 K in the experimental setup is predicted to be 3.4 mW and 4.9 mW with the original and improved heat exchangers, respectively.

Algebraic Speed Estimation for Sensorless Induction Motor Control: Insights from an Electric Vehicle Drive Cycle

Induction motors (IMs) must meet high reliability and safety standards in mission-critical applications, such as electric vehicles (EVs), where sensorless control strategies are fundamental. However, sensorless rotor speed estimation demands improvements to overcome filtering distortions, tuning complexities, and sensitivity to IM model mismatch. Algebraic methods offer inherent filtering capabilities and design flexibility to address these challenges without introducing additional dynamics into the control system. The objective of this paper is to provide an algebraic estimation strategy that yields an accurate rotor speed estimate for sensorless IM control. The strategy includes an algebraic estimator with single-parameter tuning and inherent filtering action. We propose an EV case study to experimentally evaluate and compare its performance with a typical drive cycle and a dynamic torque load that emulates a small-scale EV power train. The algebraic estimator exhibited a signal-to-noise ratio (SNR) of 43 dB. The closed-loop experiment for the EV case study showed average tracking errors below 1 rad/s and similar performance compared to a well-known sensorless strategy. Our results show that the proposed algebraic estimation strategy works effectively in a nominal speed range for a practical IM sensorless application. The algebraic estimator only requires single-parameter tuning and potentially facilitates IM model updates using a resetting scheme.

Cornerstone for MRI-compatible robots: A continuous pneumatic actuator with easy-to-connect prismatic/revolute and bidirectional encoder modules

The exceptional imaging capability of magnetic resonance imaging (MRI) enables the applications of robot-assisted surgery under MRI guidance. Essential to these robotic systems are the MRI-compatible robotic joint actuators. Among various actuation methods using MRI-compatible materials, pneumatic actuators are considered the most suitable for this application scenario due to their sterility, safety, and reliability. Yet, the current pneumatic actuators are either bulky, inefficient, or cannot achieve accurate bidirectional continuous motion. As a step forward, a novel high-performance continuous MRI-compatible pneumatic motor is proposed, along with quick-connect revolute/prismatic motion output modules and high accurate bidirectional encoder modules. The high performance of the pneumatic rotary and linear actuators are evaluated through experimental testing, where the rotary actuator can provide up to 2 Nm of output torque and over 28 W of output power, while the linear variant can deliver over 50 N of thrust, which surpass the SOTA competitors. Corresponding dynamic models are established and verified via experiments, with a maximum error of only 2.01%. Closed-loop control experiments demonstrate the high control precision and fast response frequency of the proposed pneumatic actuators, promoting the performance of joint actuators of various MRI-compatible surgical robots.

Proposal for a realtime Einstein-synchronization-defined satellite virtual clock

Abstract The realization of high performance satellite onboard clock is vital for various PNT applications. For further improvement of the synchronization-based satellite time and frequency references, this paper proposes a geosynchronous (GEO) satellite virtual clock concept based on ground–satellite synchronization and presents a beacon transponder structure for its implementation (scheduled for launch in 2025), which does not require atomic clocks to be mounted on the satellite. Its high performance relies only on minor modifications to the existing transponder structure of GEO satellites. We carefully modeled the carrier phase link and analyzed the factors causing link asymmetry within the special relativity. Considering that the performance of such synchronization-based satellite clocks is primarily limited by the link’s random phase noise, which cannot be adequately modeled, we designed a closed-loop experiment based on commercial GEO satellites for pre-evaluation. This experiment was aimed at extracting the zero-means random part of the ground-satellite Ku-band carrier phase via a feedback loop. Ultimately, we obtained a 1σ value of 0.633 ps (two-way link), following the Gaussian distribution. From this result, we conclude that the proposed real-time Einstein-synchronization-defined satellite virtual clock can achieve picosecond-level replication of onboard time and frequency.

BRAND: a platform for closed-loop experiments with deep network models.

Objective.Artificial neural networks (ANNs) are state-of-the-art tools for modeling and decoding neural activity, but deploying them in closed-loop experiments with tight timing constraints is challenging due to their limited support in existing real-time frameworks. Researchers need a platform that fully supports high-level languages for running ANNs (e.g. Python and Julia) while maintaining support for languages that are critical for low-latency data acquisition and processing (e.g. C and C++).Approach.To address these needs, we introduce the Backend for Realtime Asynchronous Neural Decoding (BRAND). BRAND comprises Linux processes, termednodes, which communicate with each other in agraphvia streams of data. Its asynchronous design allows for acquisition, control, and analysis to be executed in parallel on streams of data that may operate at different timescales. BRAND uses Redis, an in-memory database, to send data between nodes, which enables fast inter-process communication and supports 54 different programming languages. Thus, developers can easily deploy existing ANN models in BRAND with minimal implementation changes.Main results.In our tests, BRAND achieved <600 microsecond latency between processes when sending large quantities of data (1024 channels of 30 kHz neural data in 1 ms chunks). BRAND runs a brain-computer interface with a recurrent neural network (RNN) decoder with less than 8 ms of latency from neural data input to decoder prediction. In a real-world demonstration of the system, participant T11 in the BrainGate2 clinical trial (ClinicalTrials.gov Identifier: NCT00912041) performed a standard cursor control task, in which 30 kHz signal processing, RNN decoding, task control, and graphics were all executed in BRAND. This system also supports real-time inference with complex latent variable models like Latent Factor Analysis via Dynamical Systems.Significance.By providing a framework that is fast, modular, and language-agnostic, BRAND lowers the barriers to integrating the latest tools in neuroscience and machine learning into closed-loop experiments.

Gap Self-Sensing Method for Maglev System Using Partial Electromagnet Coil

The use of dedicated sensor to measure levitation gap in maglev system leads to the problems of large volume, inconvenient installation and high cost. In this article, a gap self-sensing method using partial electromagnet coil is proposed to solve these problems. The entire electromagnet coil provides levitation force, while a small part of the electromagnet coil (the subcoil) on the side facing the levitated object is injected with high frequency excitation for gap self-sensing. The theoretical derivation shows the effects of levitation gap on the impedance of the subcoil. The frequency of the injected excitation and the structure of the subcoil are determined with finite element analysis for high sensitivity. The crosstalk between the gap self-sensing system and the levitation system is suppressed by the designed signal processing based on series resonance. The prototype in this article can cover a gap measurement range of 0–20 mm for maglev ball, with a nonlinearity error of 0.2% and a measurement accuracy of 0.2 mm. The closed-loop feedback control experiment shows that the performances of the proposed gap self-sensing system meet the control requirements.

Error Compensation Method for Pedestrian Navigation System Based on Low-Cost Inertial Sensor Array.

In the pedestrian navigation system, researchers have reduced measurement errors and improved system navigation performance by fusing measurements from multiple low-cost inertial measurement unit (IMU) arrays. Unfortunately, the current data fusion methods for inertial sensor arrays ignore the system error compensation of individual IMUs and the correction of position information in the zero-velocity interval. Therefore, these methods cannot effectively reduce errors and improve accuracy. An error compensation method for pedestrian navigation systems based on a low-cost array of IMUs is proposed in this paper. The calibration method for multiple location-free IMUs is improved by using a sliding variance detector to segment the angular velocity magnitude into stationary and motion intervals, and each IMU is calibrated independently. Compensation is then applied to the velocity residuals in the zero-velocity interval after zero-velocity update (ZUPT). The experimental results show a significant improvement in the average noise performance of the calibrated IMU array, with a 3.01-fold increase in static noise performance. In the closed-loop walking experiment, the average horizontal position error of a single calibrated IMU is reduced by 27.52% compared to the uncalibrated IMU, while the calibrated IMU array shows a 2.98-fold reduction in average horizontal position error compared to a single calibrated IMU. After compensating for residual velocity, the average horizontal position error of a single IMU is reduced by 0.73 m, while that of the IMU array is reduced by 64.52%.

Ex-vivo systems for neuromodulation: A comparison of ex-vivo and in-vivo large animal nerve electrophysiology

BackgroundLittle research exists on extending ex-vivo systems to large animal nerves, and to the best of our knowledge, there has yet to be a study comparing these against in-vivo data. This paper details the first ex-vivo system for large animal peripheral nerves to be compared with in-vivo results. New MethodDetailed ex-vivo and in-vivo closed-loop neuromodulation experiments were conducted on pig ulnar nerves. Temperatures from 20 °C to 37 °C were evaluated for the ex-vivo system. The data were analysed in the time and velocity domains, and a regression analysis established how evoked compound action potential amplitude and modal conduction velocity (CV) varied with temperature and time after explantation. Main resultsPig ulnar nerves were sustained ex-vivo up to 5 h post-explantation. CV distributions of ex-vivo and in-vivo data were compared, showing closer correspondence at 37 °C. Regression analysis results also demonstrated that modal CV and time since explantation were negatively correlated, whereas modal CV and temperature were positively correlated. Comparison with Existing MethodsPrevious ex-vivo systems were primarily aimed at small animal nerves, and we are not aware of an ex-vivo system to be directly compared with in-vivo data. This new approach provides a route to understand how ex-vivo systems for large animal nerves can be developed and compared with in-vivo data. ConclusionThe proposed ex-vivo system results were compared with those seen in-vivo, providing new insights into large animal nerve activity post-explantation. Such a system is crucial for complementing in-vivo experiments, maximising collected experimental data, and accelerating neural interface development.

Research on underwater motion modeling and closed-loop control of bionic undulating fin robot

Aiming at the problems of unclear underwater motion characteristics and poor robustness of bionic undulating fin robots, the underwater motion characteristics of the robot are studied by force analysis and numerical simulation, and the dynamic model is established. Based on the dynamic model, the proportional-integral-derivative controller (PID) and neural network PID simulation model are established for numerical simulation, and the robot's underwater closed-loop motion control experiment is carried out. The accuracy of the dynamic model and the consistency between the numerical simulation and the experimental results can be verified. The results indicate that under the function of the neural network PID controller, the robot can be closer to the target depth, and the fluctuation range could be smaller. As the above characteristics improve, the control accuracy and stability of the robot could also be enhanced.

Closed-loop experiments and brain machine interfaces with multiphoton microscopy.

In the field of neuroscience, the importance of constructing closed-loop experimental systems has increased in conjunction with technological advances in measuring and controlling neural activity in live animals. We provide an overview of recent technological advances in the field, focusing on closed-loop experimental systems where multiphoton microscopy-the only method capable of recording and controlling targeted population activity of neurons at a single-cell resolution in vivo-works through real-time feedback. Specifically, we present some examples of brain machine interfaces (BMIs) using in vivo two-photon calcium imaging and discuss applications of two-photon optogenetic stimulation and adaptive optics to real-time BMIs. We also consider conditions for realizing future optical BMIs at the synaptic level, and their possible roles in understanding the computational principles of the brain.