查看更多>>摘要：Wastewater treatment processes (WWTPs) are large-scale systems comprising multiple biological reactors, which are essential for preventing water pollution and promoting water reuse. Safety assessment and accurate process monitoring are crucial for maintaining the effluent quality of WWTPs. However, the presence of uncertainties and process noise degrades the performance of fault detection models, posing significant challenges to reliable monitoring. This paper proposes a data-driven fault detection framework for monitoring failures in wastewater treatment processes affected by impulsive noise. The fault detection model employs nonlinear autoregressive with exogenous input (NARX) neural networks to construct the residual generator with the aid of robust continuous mixed p-norm optimization. Robust continuous mixed p-norm combines multiple error p-norms to enhance the cost function with diverse error information, minimizing it to produce adaptive gains that adjust the training gain based on data quality at each step. When impulsive noise occurs, the correction term for parameter estimation approaches zero, enabling the model to achieve greater robustness against impulsive noise compared to existing methods. Additionally, the fault detection model incorporates an adaptive moment estimation-based variable-step algorithm to enhance convergence by adaptively adjusting the learning rate. The proposed method is applied to the benchmark simulation model no. 1, and experimental results demonstrate that it achieves accurate detection rates for monitoring WWTPs.
查看更多>>摘要：New challenges in spatial missions and the design of new launchers entail a focus on innovative control strategies. Recent developments in Machine Learning (ML) for optimization processes shed light on the possibilities offered for controlling complex nonlinear partially unknown systems. This work focuses on the use of these methods to design control laws stabilizing the sloshing of propellants in tanks during launcher flight. A major hurdle in applying control laws designed by Artificial Intelligence (AI) to safety-critical systems lies in certifying stability and safety. Using Control Lyapunov Function (CLF) and Control Barrier Function (CBF) developed in Control Theory approaches, closed-loop stability and safety in terms of state constraints can be verified. Considering a Deep Reinforcement Learning (DRL) framework, an algorithm is developed to learn a control policy along with stability and safety certificates. The CLF and CBF conditions are integrated in the DRL algorithm, bridging the gap between Control Theory and Machine Learning techniques. A safe and stable DRL controller is then learned on a simulated launcher subject to sloshing with uncertainties and perturbations due to sloshing. A robustness study with Monte Carlo simulations is conducted to evaluate performance under various conditions. Finally, the developed controller is validated on an industrial simulator that more accurately models the real behavior of the launcher. Despite not being trained on this industrial simulator, the controller matches control objectives, demonstrating robustness and performance.
查看更多>>摘要：The unforeseen imbalance between train supply and passenger demand in the metro system usually propagates along the running direction, which increases passengers' travel costs and even seriously affects the metro system's safety and reliability. Data-driven propagation and recovery of supply-demand imbalance is thus important for a metro system. This paper proposes a two-layer imbalance propagation and recovery model for metro systems based on historical traffic data, where the affected time scope, the affected space scope, and the spatial-temporal extra waiting cost are outputted to describe the propagation and recovery mechanism. The lower-layer model calculates the passengers' latest arrival time matrix for boarding each train. This matrix is an essential input parameter of the upper-layer imbalance propagation model, where the real-time extra waiting cost for each platform under disruptions is estimated. The proposed model is applied to a real-world metro line of Shenzhen metro to analyze its spatial-temporal propagation and recovery characteristics facing imbalance, which is of notable significance to the possible performance optimization and safety assessment for the metro system.
查看更多>>摘要：Drilling tools loading and unloading for drilling machine is a labor-intensive task. The automatic loading and unloading function of the drilling machine can prominently lessen the labor intensity borne by workers and serves as a key requisite for the full automation of underground drilling equipment. The drilling tools loading and unloading system (DTLUS) is subject to a relatively large load, leading to fluctuations in the hydraulic system's pressure and flow rate. In view of this, a robust control strategy based on the sliding mode controller is proposed. Aiming at the requirement of suppressing chattering during the operation of the sliding mode controller (SMC), a fuzzy neural network (FNN) parameter adjustment method based on the variable structure control framework is designed, which is intended to enhance the stability and control accuracy of the system, so as to better cope with the complex working conditions of the DTLUS and ensure its efficient and stable operation. The experimental results indicate that the designed controller can reduce mechanical impact and enhance the efficiency of drilling tools loading and unloading.
查看更多>>摘要：An industrial alarm flood indicates the emergence of a major problem and requires immediate and effective measures to mitigate the situation. Without essential information on the current and upcoming alarms, it is difficult to respond efficiently, especially when the alarm rate is significantly high. Thus, a reinforcement learning (RL) approach is proposed in this work for early prediction of an industrial alarm flood so as to provide critical decision support to industrial operators in real-time. The proposed method is implemented mainly in the following steps: (1) An alarm flood pattern extraction strategy is adopted to exclude irrelevant alarms and generate potential online scenarios by exploiting the alarm relations in the historical alarm flood sequences; (2) To analyze alarm information effectively, a novel textual vectorization method based on mutual information is proposed; (3) Finally, the early prediction problem is formulated as a partially observable Markov decision process (POMDP) and the double deep Q-network (DDQN) algorithm is adopted, with modifications to the learning process to ensure both accuracy and earliness. The effectiveness of the proposed method is demonstrated using real industrial data from an oil refinery.
查看更多>>摘要：When an optical telescope tracks dim celestial bodies, the camera requires a long-time exposure to obtain clear target images. Due to the self-rotation of the Earth, for an alt-azimuth telescope, this will result in a rotation of the field of view. To counteract the rotation, an eliminating image rotation mechanism (EIRM) is required. This paper introduces a new EIRM, which is low cost and easy to install. Due to the unavoidable backlash and nonlinearity of stepper motor, the accurate model of EIRM is difficult to obtain. In order to improve the control accuracy of EIRM, a novel control algorithm based on sigmoid function and speed feedforward is proposed. Control experiments based on proportional integral (PI) control, Fuzzy-PI and proposed controller were conducted, and the results showed that the proposed algorithm has the smallest tracking error. Lastly, the mechanism is installed on a 1.2-meter aperture optical telescope, and the observation results further demonstrate the effectiveness of this method.
查看更多>>摘要：This paper proposes an autonomous steering control framework for Distributed Drive Autonomous Electric Vehicles (DDAEV). This framework aims to enhance trajectory tracking accuracy and vehicle stability in challenging road conditions, such as rain and snow. The methodology begins with the design of a tire slip angle estimation strategy, which utilizes a 2-DoF single-track vehicle model and a sliding mode observer to account for tire cornering characteristics. Next, a lateral stability control strategy that incorporates tire slip angle considerations is developed based on sliding mode control (SMC) and lateral stability analysis. Additionally, a trajectory tracking control strategy is proposed, integrating a dual-motor autonomous steering system. This system combines model predictive control (MPC) and a steering rack displacement tracking controller to achieve accurate tracking of the target trajectory. Finally, simulation and Hardware-in-the-Loop (HiL) test results demonstrate that the proposed control framework for DDAEVs significantly enhances trajectory tracking accuracy and stability under wet road surface.
查看更多>>摘要：A key challenge of steer-by-wire systems is that they require a desired steering angle to be followed accurately. In this paper, a steering tracking control method with prescribed performance under sensor failure is proposed to enhance the tracking performance and reliability of steering execution. To mitigate the adverse effect of the coupled steering resistance moment and estimate the unmeasured state variable, a rack force estimator based on an extended disturbance observer is developed. By designing a new preset time performance function, the tracking error is guaranteed to converge to a predefined quantitative constraint within a preset time. Notably, the convergence time and the tracking accuracy can be set arbitrarily, independent of system parameters and initial states. The present study presents a fast fault detection method and formulated corresponding fault-tolerant strategies, aiming to ensure the continued proper functioning of steering tracking control in the event of sensor failures. Through verification and comparison with hardware-in-the-loop experiments, the proposed method achieves high transient and high-precision tracking performance even under aggressive steering and sensor failure conditions, while being easily embedded applied in engineering practice.
查看更多>>摘要：This article presents multivariable control designs to improve performance and reduce fatigue loads in wind turbines using torque control, Collective Pitch Control (CPC), and Individual Pitch Control (IPC). Two multi-input, multi-output (MIMO) control structures are proposed in this study. The first approach uses force distribution on blades to design the IPC, while the second uses the Relative Gain Array (RGA), which quantifies the level of interactions between inputs and outputs, to design the IPC. Both approaches use the same torque control and CPC. This work emphasizes the second approach, i.e., the RGA-based IPC. A novel aspect of this approach is the dynamic allocation of IPC, which refers to the change in the input-output pairing as a function of the rotor azimuth angle. The frequency-dependent characteristics of this allocation facilitate fatigue load reduction at targeted frequencies. Extensive simulations show that the RGA-based IPC controller outperforms the first controller in reducing cyclic loads on the blade root bending moment, tower side-to-side, and tower fore-aft bending moments at the frequencies of interest. Moreover, it has no detrimental effects on the rotor speed and power generation, which are regulated by the CPC and torque controller. A Control-oriented, Reconfigurable, and Acausal Floating Turbine Simulator (CRAFTS), developed in-house, is used for design, implementation, and evaluation.
NETL
NSTL
Elsevier