Real-time transmission line switching control is a cost-effective and efficient measure to alleviate line overload. However, due to the nonlinear characteristics of the transmission network structure optimization prob...
详细信息
ISBN:
(纸本)9798350375145;9798350375138
Real-time transmission line switching control is a cost-effective and efficient measure to alleviate line overload. However, due to the nonlinear characteristics of the transmission network structure optimization problem, conventional optimization methods often struggle to find solutions quickly. In recent years, data-driven methods have been widely applied in fields such as power system dispatch optimization. This paper proposes a method based on deep neural networks. By using historical and simulation data of the power system, the neural network is trained to perform exploratory learning and find optimal topology control strategies when line overload occurs. This method is trained and tested on the ieee 14-bus open-source dataset, and the test results verify its effectiveness.
Many application fields, e.g., robotic surgery, autonomous piloting, and wearable robotics greatly benefit from advances in robotics and automation. A common task is to control an unknown nonlinear system such that it...
详细信息
Many application fields, e.g., robotic surgery, autonomous piloting, and wearable robotics greatly benefit from advances in robotics and automation. A common task is to control an unknown nonlinear system such that its output tracks a desired reference signal for a finite duration of time. A learningcontrol method that automatically and efficiently designs output feedback controllers for this task would greatly boost practicality over time-consuming and labour-intensive manual system identification and controller design methods. In this contribution we propose Automatic Neural Ordinary Differential Equation control (ANODEC), a data-efficient automatic design of output feedback controllers for finite-time reference tracking in systems with unknown nonlinear dynamics. In a two-step approach, ANODEC first identifies a neural ODE model of the system dynamics from input-output data of the system dynamics and then exploits this data-driven model to learn a neural ODE feedback controller, while requiring no knowledge of the actual system state or its dimensionality. In-silico validation shows that ANODEC is able to -automatically- design competitive controllers that outperform two controller baselines, and achieves an on average & AP;30 % / 17 % lower median RMSE. This is demonstrated in four different nonlinear systems using multiple, qualitatively different and even out-of-training-distribution reference signals.
In robust optimization for power system operations, striking a balance between solution robustness and performance is crucial. Unlike conventional interval-based uncertainty sets, which treat random variables as indep...
详细信息
In robust optimization for power system operations, striking a balance between solution robustness and performance is crucial. Unlike conventional interval-based uncertainty sets, which treat random variables as independent entities, our approach introduces a compact, coupled representation of these variables. We establish theoretical benchmarks to assess the benefits of employing this coupled uncertainty set in the context of the economic dispatch problem. Moreover, we have devised a pioneering data-driven algorithm capable of autonomously learning the shape of the parametric uncertainty set. This algorithm concurrently optimizes performance and furnishes solutions with statistical guarantees in terms of generalization capabilities. The effectiveness of this algorithm is validated through case studies on both a synthetic dataset and a real-world problem.
This paper introduces an intelligent optimization framework that integrates Digital Twin (DT) technology, deep learning, and a tailored Multi-Restart Bayesian Optimization with Random Initialization (MRBORI) to enhanc...
详细信息
This paper introduces an intelligent optimization framework that integrates Digital Twin (DT) technology, deep learning, and a tailored Multi-Restart Bayesian Optimization with Random Initialization (MRBORI) to enhance parameter control and yield in semiconductor manufacturing. The proposed framework synergizes XGBoost-based feature selection, which identifies critical parameters in high-dimensional spaces, with a custom deep learning surrogate model that captures complex nonlinear interactions. Building on these insights, the MRBORI strategy leverages multiple optimization restarts, each initialized randomly, to mitigate local minima risks and systematically explore broad parameter spaces. Experimental validation using real-world data from an epitaxial silicon carbide (Epi SiC) process demonstrates notably tighter thickness control and improved yield compared to traditional methods. By unifying DT-driven real-time insights with advanced machine learning and multi-restart optimization, this framework offers a robust and precise solution for tackling the complexities of modern semiconductor manufacturing.
This article delves into designing stabilizing feedback control gains for continuous-time linear systems with unknown state matrix, in which the control gain is subjected to a structural constraint. We bring forth the...
详细信息
This article delves into designing stabilizing feedback control gains for continuous-time linear systems with unknown state matrix, in which the control gain is subjected to a structural constraint. We bring forth the ideas from reinforcement learning (RL) in conjunction with sufficient stability and performance guarantees in order to design these structured gains using the trajectory measurements of states and controls. We first formulate a model-based linear quadratic regulator (LQR) framework to compute the structured control gain. Subsequently, we transform this model-based LQR formulation into a data-driven RL algorithm to remove the need for knowing the system state matrix. Theoretical guarantees are provided for the stability of the closed-loop system and the convergence of the structured RL (SRL) algorithm. A remarkable application of the proposed SRL framework is in designing distributed static feedback control, which is necessary for automatic control of many large-scale cyber-physical systems. As such, we validate our theoretical results with numerical simulations on a multiagent networked linear time-invariant dynamical system.
This article introduces the control and operation of a grid-connected converter with an energy storage system. A complete mathematical model was presented for the developed converter and its control system. The system...
详细信息
This article introduces the control and operation of a grid-connected converter with an energy storage system. A complete mathematical model was presented for the developed converter and its control system. The system under study was a small microgrid comprising an AC grid that is feeding a DC load through a converter. The converter was connected to the AC grid through an R-L filter. The classical linear controllers have limitations due to their slow transient performance and low robustness against parameter variations and load disturbances. In this paper, machine-learned controllers were used to dealing with those drawbacks of the traditional controller. First, a study for conventional nested loop Proportional Integral (PI) was introduced for both outer and inner loops PI-PI controller. A data-driven Online learning (DDOL) controller was then proposed. A comparison between the normal traditional PI-PI controller and the proposed DDOL ones was made under different operating scenarios. The converter control was tested under various operational conditions, and its dynamic and steady-state behavior was analyzed. The model was done through a MATLAB Simulink to check the normal operation of the network in a grid-connected mode under different load disturbances and AC input voltage. Then, the system was designed, fabricated, and implemented in a hardware environment in our Energy systems Research Laboratory (ESRL) testbed, and the hardware test results were verified. The results showed that the proposed DDOL controller was more robust and had better transient and steady state performances.
Model predictive control (MPC) is widely used in industries but implementing it poses challenges due to hardware or time constraints. A promising solution is to approximate the MPC policy using function approximators ...
详细信息
Model predictive control (MPC) is widely used in industries but implementing it poses challenges due to hardware or time constraints. A promising solution is to approximate the MPC policy using function approximators like neural networks. Existing methods focus on minimizing the error between the approximators outputs and the MPC optimal control actions on training data, which is called error guided learning approach in this letter. However, the goals of control law design is not to minimize the fitting error but to minimize the operation cost. This letter proposes a novel cost-guided learning approach that utilizes the cost sensitivity information from the MPC problem to directly minimize the loss in closed-loop performance. A theoretical analysis shows cost-guided learning provides tighter guarantees on optimality loss compared to traditional error-guided learning. Experiments on a continuous stirred tank reactor (CSTR) benchmark demonstrate that the proposed technique results in approximate MPC policies that achieve substantially better closed-loop performance. This letter makes an important contribution by connecting the fitting errors with operational objectives, overcoming key limitations of existing approximation methods. The core idea could be applied more broadly for data-drivencontrol.
The gap metric is a very essential tool for robust stability analysis of controlsystems and fault tolerant control. It is widely calculated as a model base. However, it is challenging to derive a precise mathematical...
详细信息
In recent years, deep learning techniques have been widely applied in soft sensor modeling. Stacked autoencoder (SAE) networks are particularly effective at discovering complex data patterns due to their hierarchical ...
详细信息
ISBN:
(纸本)9798350321050
In recent years, deep learning techniques have been widely applied in soft sensor modeling. Stacked autoencoder (SAE) networks are particularly effective at discovering complex data patterns due to their hierarchical structures. However, process data are typically generated as data streams, which poses a great challenge to capture the time-varying characteristics of the process for traditional soft sensor models based on SAE. Furthermore, the insufficiency of offline pre-training data further limits the feature representation capability of SAE. To address these problems, an online deep evolving fuzzy system (ODEFS) based adaptive soft sensor method for process data streams is proposed. In the offline modeling phase, quality-related stacked autoencoder (QSAE) is pre-trained as representation layer to mine quality-related feature representations, while an evolving fuzzy system with self-organization capability is built as the prediction layer. In the online implementation phase, the topology-preserving loss is added to the learning process of QSAE feature network to enable continuous learning of feature representations and alleviate the catastrophic forgetting problem. Meanwhile, the shallow EFS network handles concept drift in data patterns by self-adjusting the structure and parameters. The proposed ODEFS method can improve the feature representation capability of SAE in a data streaming environment and the ability to handle time-varying characteristics, thus ensuring better prediction accuracy. The effectiveness and superiority of the proposed method are verified on TE process.
Reinforcement learning (RL), induced by the exploration-exploitation dilemma, usually requires enough interaction experiences with the environment to improve performance. Rule-based methods utilize the internal expert...
详细信息
暂无评论