Adaptive Fault-Tolerant Control for Nonlinear Systems With Multiple Sensor Faults and Unknown Control Directions

Prescribed Performance Bounded-H∞ Control for Flexible-Joint Manipulators Without Initial Condition Restriction

Flexible-joint manipulators have a lightweight nature, compact structure, and high flexibility, making them widely applicable in industrial manufacturing, biomedical instruments, and aerospace fields. However, the inherent flexibility of single-link flexible-joint manipulators (SLFJMs) poses substantial control challenges. Compared to traditional control algorithms, prescribed performance control (PPC) algorithms provide superior transient response and steady-state performance by defining a prescribed performance function. However, existing PPC algorithms are limited to a specific range of system initial states, which reduces the joint manipulator's operational workspace and weakens the robustness of the control algorithm. To address this issue, this study proposes a prescribed performance bounded-H∞ fault-tolerant controller for SLFJMs. By designing an improved tangent-type barrier Lyapunov function (BLF), a prescribed performance controller that is independent of the initial state of the SLFJM is developed. An input control function (ICF) is employed to mitigate the impulse response of the control input, ensuring a smooth transition from zero. Furthermore, the improved tangent-type BLF enables the tracking error to rapidly converge to a small neighborhood of zero. Finally, a stabilization control simulation experiment is conducted; the results validate the effectiveness of the proposed prescribed performance bounded-H∞ controller.

Observer-Based Adaptive Fixed-Time Sensor Fault Compensation Control for Uncertain Nonlinear Systems

This article focuses on an observer-based adaptive sensor fault compensation fixed-time tracking control problem for uncertain nonlinear systems. A sixth-power Lyapunov function is designed for the first time which lays the foundation to construct the effective adaptive fixed-time fault compensation mechanism. Meanwhile, in the controller design procedure, owing to the existence of the sensor fault, only the actual output can be measured, unlike existing results, an improved state observer is constructed to estimate the unmeasured states effectively. Under our developed control mechanism, all closed-loop signals are bounded within fixed-time interval, observation errors and tracking error can converge into a small domain around zero. Simulation verifies the availability of the presented approach further.

Adaptive output feedback fault-tolerant control for a class of nonlinear systems based on a sensor fusion mechanism

This paper investigates an output feedback adaptive fault-tolerant tracking control for a class of nonlinear systems with system nonlinearities, sensor failures and external disturbances, in which sensor redundancy is employed to enhance measurement reliability. A sensor fusion mechanism, together with a novel history-based weighted average algorithm is first designed to fuse all sensor outputs. Then, an adaptive controller based on the sensor fusion output, a dynamic gain and a state observer is constructed to handle all the uncertainties caused by system nonlinearities, external disturbances, sensor failures and fusion mechanism. It is shown that by using the proposed scheme, the closed-loop system is stable, the sensor fusion mechanism can eliminate the effects of faulty sensors, and the real tracking error can be driven into a small compact set mainly affected by the fusion error. Experimental results are accomplished to validate the proposed scheme.

Cooperative-Critic Learning-Based Secure Tracking Control for Unknown Nonlinear Systems With Multisensor Faults

This article develops a cooperative-critic learning-based secure tracking control (CLSTC) method for unknown nonlinear systems in the presence of multisensor faults. By introducing a low-pass filter, the sensor faults are transformed into "pseudo" actuator faults, and an augmented system that integrates the system state and the filter output is constructed. To reduce design costs, a joint neural network Luenberger observer (NNLO) structure is established by using neural network and input/output data of the system to identify unknown system dynamics and sensor faults online. To achieve the optimal secure tracking control, an augmented tracking system is formed by integrating the dynamics of tracking error, reference trajectory, and filter output. Then, a novel cost function is designed for the augmented tracking system, which employs the fault estimation and the discount factor. The Hamilton-Jacobi-Bellman equation is solved to obtain the CLSTC strategy through an adaptive critic structure with cooperative tuning laws. Besides, the Lyapunov stability theorem is utilized to prove that all signals of the closed-loop system converge to a small neighborhood of the equilibrium point. Simulation results demonstrate that the proposed control method has good fault tolerance performance and is suitable for solving secure control problems of nonlinear systems with various sensor faults.

Machine Learning Molecular Dynamics Shows Anomalous Entropic Effect on Catalysis through Surface Pre-melting of Nanoclusters

Due to the superior catalytic activity and efficient utilization of noble metals, nanocatalysts are extensively used in the modern industrial production of chemicals. The surface structures of these materials are significantly influenced by reactive adsorbates, leading to dynamic behavior under experimental conditions. The dynamic nature poses significant challenges in studying the structure-activity relations of catalysts. Herein, we unveil an anomalous entropic effect on catalysis via surface pre-melting of nanoclusters through machine learning accelerated molecular dynamics and free energy calculation. We find that due to the pre-melting of shell atoms, there exists a non-linear variation in the catalytic activity of the nanoclusters with temperature. Consequently, two notable changes in catalyst activity occur at the respective temperatures of melting for the shell and core atoms. We further study the nanoclusters with surface point defects, i.e. vacancy and ad-atom, and observe significant decrease in the surface melting temperatures of the nanoclusters, enabling the reaction to take place under more favorable and milder conditions. These findings not only provide novel insights into dynamic catalysis of nanoclusters but also offer new understanding of the role of point defects in catalytic processes.

Prescribed Performance Fault-Tolerant Control of Strict-Feedback Systems via Error Shifting

This article investigates the prescribed performance control (PPC) problem for a class of nonlinear strict-feedback systems with sensor/actuator faults. A shifting function is introduced to modify the output tracking error generated by the practically measured system state, based on which an improved PPC method is proposed to achieve the convergence of output tracking error to the prescribed region, and this convergence is shown to be independent of the initial tracking condition and insusceptible to sensor/actuator faults. The faults-induced uncertainties together with the nonlinear dynamics are compensated by involving a radial basis function neural network (RBFNN) to make the controller robust adaptive fault-tolerant without prior knowledge of fault coefficients. Via Lyapunov stability analysis, it is proven that all signals in the closed-loop system are semiglobally uniformly ultimately bounded. The effectiveness and superiority of the method are demonstrated by two simulation examples.

Neuroadaptive Fault-Tolerant Control With Guaranteed Performance for Euler-Lagrange Systems Under Dying Power Faults

This article investigates the tracking control problem for Euler-Lagrange (EL) systems subject to output constraints and extreme actuation/propulsion failures. The goal here is to design a neural network (NN)-based controller capable of guaranteeing satisfactory tracking control performance even if some of the actuators completely fail to work. This is achieved by introducing a novel fault function and rate function such that, with which the original tracking control problem is converted into a stabilization one. It is shown that the tracking error is ensured to converge to a pre-specified compact set within a given finite time and the decay rate of the tracking error can be user-designed in advance. The extreme actuation faults and the standby actuator handover time delay are explicitly addressed, and the closed signals are ensured to be globally uniformly ultimately bounded. The effectiveness of the proposed method has been confirmed through both theoretical analysis and numerical simulation.

Bridging semiempirical and ab initio QM/MM potentials by Gaussian process regression and its sparse variants for free energy simulation

Free energy simulations that employ combined quantum mechanical and molecular mechanical (QM/MM) potentials at ab initio QM (AI) levels are computationally highly demanding. Here, we present a machine-learning-facilitated approach for obtaining AI/MM-quality free energy profiles at the cost of efficient semiempirical QM/MM (SE/MM) methods. Specifically, we use Gaussian process regression (GPR) to learn the potential energy corrections needed for an SE/MM level to match an AI/MM target along the minimum free energy path (MFEP). Force modification using gradients of the GPR potential allows us to improve configurational sampling and update the MFEP. To adaptively train our model, we further employ the sparse variational GP (SVGP) and streaming sparse GPR (SSGPR) methods, which efficiently incorporate previous sample information without significantly increasing the training data size. We applied the QM-(SS)GPR/MM method to the solution-phase SN2 Menshutkin reaction, NH3+CH3Cl→CH3NH3++Cl-, using AM1/MM and B3LYP/6-31+G(d,p)/MM as the base and target levels, respectively. For 4000 configurations sampled along the MFEP, the iteratively optimized AM1-SSGPR-4/MM model reduces the energy error in AM1/MM from 18.2 to 4.4 kcal/mol. Although not explicitly fitting forces, our method also reduces the key internal force errors from 25.5 to 11.1 kcal/mol/Å and from 30.2 to 10.3 kcal/mol/Å for the N-C and C-Cl bonds, respectively. Compared to the uncorrected simulations, the AM1-SSGPR-4/MM method lowers the predicted free energy barrier from 28.7 to 11.7 kcal/mol and decreases the reaction free energy from -12.4 to -41.9 kcal/mol, bringing these results into closer agreement with their AI/MM and experimental benchmarks.

Data-Driven Formation Control for Unknown MIMO Nonlinear Discrete-Time Multi-Agent Systems With Sensor Fault

A data-driven distributed formation control algorithm is proposed for an unknown heterogeneous non-affine nonlinear discrete-time MIMO multi-agent system (MAS) with sensor fault. For the considered unknown MAS, the dynamic linearization technique in model-free adaptive control (MFAC) theory is used to transform the unknown MAS into an equivalent virtual dynamic linearization data model. Then using the virtual data model, the structure of the distributed model-free adaptive controller is constructed. For the incorrect signal measurements due to the sensor fault, the radial basis function neural network (RBFNN) is first trained for the MAS under the fault-free case, then using the outputs of the well-trained RBFNN and the actual outputs of MAS under sensor fault case, the estimation laws of the unknown fault and system parameters in the virtual data model are designed with only the measured input-output (I/O) data information. Finally, the boundedness of the formation error is analyzed by the contraction mapping method and mathematical induction method. The effectiveness of the proposed algorithm is illustrated by simulation examples.

Rapid Sensor Fault Diagnosis for a Class of Nonlinear Systems via Deterministic Learning

In this article, a rapid sensor fault diagnosis (SFD) method is presented for a class of nonlinear systems. First, by exploiting the linear adaptive observer technology and the deterministic learning method (DLM), an adaptive neural network (NN) observer is constructed to capture the information of the unknown sensor fault function. Second, when the NN input orbit is a period or recurrent one, the partial persistent excitation (PE) condition of the NNs can be guaranteed through the DLM. Based on the partial PE condition and the uniformly completely observable property of a linear time-varying system, the accurate state estimation and the sensor fault identification can be achieved by properly choosing the observer gain. Third, a bank of dynamical observers utilizing the experiential knowledge is constructed to achieve rapid SFD and data recovery. The attractions of the proposed approach are that accurate approximations of sensor faults can be achieved through the DLM, and the data that are destroyed by the sensor faults can be recovered by using the learning results. Simulation studies of a robot system are utilized to show the effectiveness of the proposed method.

Event-Triggered Adaptive Neural Network Sensor Failure Compensation for Switched Interconnected Nonlinear Systems With Unknown Control Coefficients

In this article, a decentralized adaptive neural network (NN) event-triggered sensor failure compensation control issue is investigated for nonlinear switched large-scale systems. Due to the presence of unknown control coefficients, output interactions, sensor faults, and arbitrary switchings, previous works cannot solve the investigated issue. First, to estimate unmeasured states, a novel observer is designed. Then, NNs are utilized for identifying both interconnected terms and unstructured uncertainties. A novel fault compensation mechanism is proposed to circumvent the obstacle caused by sensor faults, and a Nussbaum-type function is introduced to tackle unknown control coefficients. A novel switching threshold strategy is developed to balance communication constraints and system performance. Based on the common Lyapunov function (CLF) method, an event-triggered decentralized control scheme is proposed to guarantee that all closed-loop signals are bounded even if sensors undergo failures. It is shown that the Zeno behavior is avoided. Finally, simulation results are presented to show the validity of the proposed strategy.

Active Fault-Tolerant Control for Near-Space Hypersonic Vehicles

Due to the harsh working environment, Near-Space Hypersonic Vehicles (NSHVs) have the characteristics of frequent faults, which seriously affect flight safety. However, most researches focus on active fault-tolerant control for actuator faults. In order to fill the gap of active fault-tolerant control for sensor faults, this paper presents an Active Fault-Tolerant Control (AFTC) strategy for NSHVs based on Active Disturbance Rejection Control (ADRC) combined with fault diagnosis and evaluation. With the proposed AFTC strategy, both sensor faults and actuator faults can be compensated within 0.5 s. Wavelet packet decomposition and Kernel Extreme Learning Machine (KELM) are associated to ensure the high accuracy and real-time ability of fault diagnosis. Simulation results show that the proposed fault diagnosis method can significantly reduce the divergence of diagnosis results by up to 98%. The fault information is used to generate tolerant compensation, which is combined with the ADRC to achieve AFTC. Statistical results indicate that AFTC has significantly lower static error than ADRC. The proposed AFTC method endows NSHVs with the ability to complete missions even when various types of faults appear. Its advantages are demonstrated in comparison with other fault diagnosis and tolerant control methods.

Enabling Large-Scale Condensed-Phase Hybrid Density Functional Theory-Based Ab Initio Molecular Dynamics II: Extensions to the Isobaric-Isoenthalpic and Isobaric-Isothermal Ensembles

In the previous paper of this series [Ko, H.-Y. et al. J. Chem. Theory Comput. 2020, 16, 3757-3785], we presented a theoretical and algorithmic framework based on a localized representation of the occupied space that exploits the inherent sparsity in the real-space evaluation of the exact exchange (EXX) interaction in finite-gap systems. This was accompanied by a detailed description of exx, a massively parallel hybrid message-passing interface MPI/OpenMP implementation of this approach in Quantum ESPRESSO (QE) that enables linear scaling hybrid density functional theory (DFT)-based ab initio molecular dynamics (AIMD) in the microcanonical/canonical (NVE/NVT) ensembles of condensed-phase systems containing 500-1000 atoms (in fixed orthorhombic cells) with a wall time cost comparable to semi-local DFT. In this work, we extend the current capabilities of exx to enable hybrid DFT-based AIMD simulations of large-scale condensed-phase systems with general and fluctuating cells in the isobaric-isoenthalpic/isobaric-isothermal (NpH/NpT) ensembles. The theoretical extensions to this approach include an analytical derivation of the EXX contribution to the stress tensor for systems in general simulation cells with a computational complexity that scales linearly with system size. The corresponding algorithmic extensions to exx include optimized routines that (i) handle both static and fluctuating simulation cells with non-orthogonal lattice symmetries, (ii) solve Poisson's equation in general/non-orthogonal cells via an automated selection of the auxiliary grid directions in the Natan-Kronik representation of the discrete Laplacian operator, and (iii) evaluate the EXX contribution to the stress tensor. Using this approach, we perform a case study on a variety of condensed-phase systems (including liquid water, a benzene molecular crystal polymorph, and semi-conducting crystalline silicon) and demonstrate that the EXX contributions to the energy and stress tensor simultaneously converge with an appropriate choice of exx parameters. This is followed by a critical assessment of the computational performance of the extended exx module across several different high-performance computing architectures via case studies on (i) the computational complexity due to lattice symmetry during NpT simulations of three different ice polymorphs (i.e., ice Ih, II, and III) and (ii) the strong/weak parallel scaling during large-scale NpT simulations of liquid water. We demonstrate that the robust and highly scalable implementation of this approach in the extended exx module is capable of evaluating the EXX contribution to the stress tensor with negligible cost (<1%) as well as all other EXX-related quantities needed during NpT simulations of liquid water (with a very tight 150 Ry planewave cutoff) in ≈5.2 s ((H₂O)₁₂₈) and ≈6.8 s ((H₂O)₂₅₆) per AIMD step. As such, the extended exx module presented in this work brings us another step closer to routinely performing hybrid DFT-based AIMD simulations of sufficient duration for large-scale condensed-phase systems across a wide range of thermodynamic conditions.

Development of Range-Corrected Deep Learning Potentials for Fast, Accurate Quantum Mechanical/Molecular Mechanical Simulations of Chemical Reactions in Solution

We develop a new deep potential─range correction (DPRc) machine learning potential for combined quantum mechanical/molecular mechanical (QM/MM) simulations of chemical reactions in the condensed phase. The new range correction enables short-ranged QM/MM interactions to be tuned for higher accuracy, and the correction smoothly vanishes within a specified cutoff. We further develop an active learning procedure for robust neural network training. We test the DPRc model and training procedure against a series of six nonenzymatic phosphoryl transfer reactions in solution that are important in mechanistic studies of RNA-cleaving enzymes. Specifically, we apply DPRc corrections to a base QM model and test its ability to reproduce free-energy profiles generated from a target QM model. We perform these comparisons using the MNDO/d and DFTB2 semiempirical models because they differ in the way they treat orbital orthogonalization and electrostatics and produce free-energy profiles which differ significantly from each other, thereby providing us a rigorous stress test for the DPRc model and training procedure. The comparisons show that accurate reproduction of the free-energy profiles requires correction of the QM/MM interactions out to 6 Å. We further find that the model's initial training benefits from generating data from temperature replica exchange simulations and including high-temperature configurations into the fitting procedure, so the resulting models are trained to properly avoid high-energy regions. A single DPRc model was trained to reproduce four different reactions and yielded good agreement with the free-energy profiles made from the target QM/MM simulations. The DPRc model was further demonstrated to be transferable to 2D free-energy surfaces and 1D free-energy profiles that were not explicitly considered in the training. Examination of the computational performance of the DPRc model showed that it was fairly slow when run on CPUs but was sped up almost 100-fold when using NVIDIA V100 GPUs, resulting in almost negligible overhead. The new DPRc model and training procedure provide a potentially powerful new tool for the creation of next-generation QM/MM potentials for a wide spectrum of free-energy applications ranging from drug discovery to enzyme design.

Performance-Guaranteed Fault-Tolerant Control for Uncertain Nonlinear Systems via Learning-Based Switching Scheme

This article is concerned with the challenge of guaranteeing output constraints for fault-tolerant control (FTC) of a class of unknown multi-input single-output (MISO) nonlinear systems in the presence of actuator faults. Most industrial systems are equipped with redundant actuators and a fault detection-isolation mechanism for accommodating unexpected actuator faults. To simplify the system design and reduce the risk of false alarm or missed detection brought by the detection unit, a learning-based switching function scheme is proposed to automatically activate different sets of actuators in a rotational manner without human intervention. By this means, no explicit fault detection mechanism is needed. An additional step has been made to guarantee that the system output remains in user-defined time-varying asymmetric output constraints all the time during the occurrence of failures by utilizing error transformation techniques. The stability of the transformed system can equivalently deliver the result that the original system output stays in the required bounds. Hence, system crash or further catastrophic outcomes can be avoided. A neural network is integrated to embody the adaptive FTC design for dealing with unknown system dynamics. The dynamic surface control (DSC) technique is also invoked to decrease complexity. Furthermore, the stability analysis is carried out by the standard Lyapunov approach to guarantee that all the signals of the closed-loop system are semiglobally uniformly ultimately bounded. Finally, the simulation results are provided to verify the effectiveness of the proposed scheme.

Adaptive Sliding Mode Control for a Robotic Manipulator with Unknown Friction and Unknown Control Direction

This paper is aimed at addressing the tracking control issue for an n-DOF manipulator regardless of unknown friction and unknown control direction. In order to handle the above issues, an adaptive sliding mode control (ASMC) is developed with a Nussbaum function. The sliding mode control (SMC) in the proposed control guarantees the tracking problem and fast responses for the manipulator. Additionally, there are adaptive laws for the robust gain in the SMC to deal with the unknown external disturbance and reduce the chattering effect in the system. In practice, the mistakes in the connection between actuators and drivers, named unknown control direction, cause serious damage to the manipulator. To overcome this issue, the Nussbaum function is multiplied by the ASMC law. A Lyapunov approach is investigated to analyze the stability and robustness of the whole system. Finally, several simulations are implemented on a 3-DOF manipulator and their results are compared with those of the existing controllers to validate the advantages of the proposed method.

Adaptive Optimal Control for a Class of Nonlinear Systems: The Online Policy Iteration Approach

This paper studies the online adaptive optimal controller design for a class of nonlinear systems through a novel policy iteration (PI) algorithm. By using the technique of neural network linear differential inclusion (LDI) to linearize the nonlinear terms in each iteration, the optimal law for controller design can be solved through the relevant algebraic Riccati equation (ARE) without using the system internal parameters. Based on PI approach, the adaptive optimal control algorithm is developed with the online linearization and the two-step iteration, i.e., policy evaluation and policy improvement. The convergence of the proposed PI algorithm is also proved. Finally, two numerical examples are given to illustrate the effectiveness and applicability of the proposed method.

Adaptive sensor fault-tolerant control for a class of multi-input multi-output nonlinear systems: Adaptive first-order filter-based dynamic surface control approach

This paper is concerned with the adaptive fault-tolerant control (FTC) problem for a class of multivariable nonlinear systems with external disturbances, modeling errors and time-varying sensor faults. The bias, drift, loss of accuracy and loss of effectiveness faults can be effectively accommodated by this scheme. The dynamic surface control (DSC) technique and adaptive first-order filters are brought together to design an adaptive FTC scheme which can reduce significantly the computational burden and improve further the control performance. The adaptation laws are constructed using novel low-pass filter based modification terms which enable under high learning or modification gains to achieve robust, fast and high-accuracy estimation without incurring undesired high-frequency oscillations. It is proved that all signals in the closed-loop system are uniformly ultimately bounded and the tracking-errors can be made arbitrary close to zero. Simulation results are provided to verify the effectiveness and superiority of the proposed FTC method.