检索结果-内蒙古大学图书馆

A Unified Framework for Online Data-Driven Predictive control with Robust Safety Guarantees

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Vahidi-Moghaddam, Amin Chen, Kaian Zhang, Kaixiang Li, Zhaojian Wang, Yan Wu, Kai The Department of Mechanical Engineering Michigan State University East LansingMI48824 United States The Research and Advanced Engineering Ford Motor Company DearbornMI48121 United States

Despite great successes, model predictive control (MPC) relies on an accurate dynamical model and requires high onboard computational power, impeding its wider adoption in engineering systems, especially for nonlinear real-time systems with limited computation power. These shortcomings of MPC motivate this work to make such a control framework more practically viable for real-world applications. Specifically, to remove the required accurate dynamical model and reduce the computational cost for nonlinear MPC (NMPC), this paper develops a unified online data-driven predictive control pipeline to efficiently control a system with guaranteed safety without incurring large computational complexity. The new aspect of this idea is learning not only the real system but also the control policy, which results in a reasonable computational cost for the data-driven predictive controllers. More specifically, we first develop a spatial temporal filter (STF)-based concurrent learning scheme to systematically identify system dynamics for general nonlinear systems. We then develop a robust control barrier function (RCBF) for safety guarantees in the presence of model uncertainties and learn the RCBF-based NMPC policy. Furthermore, to mitigate the performance degradation due to the existing model uncertainties, we propose an online policy correction scheme through perturbation analysis and design of an ancillary feedback controller. Finally, extensive simulations on two applications, cart-inverted pendulum and automotive powertrain control, are performed to demonstrate the efficacy of the proposed framework, which shows comparable performance with much lower computational cost in comparison with several benchmark algorithms. © 2023, CC0.

关键词： Model predictive control

KARNet: Kalman Filter Augmented Recurrent Neural Network for Learning World Models in Autonomous Driving Tasks

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Manjunatha, Hemanth Pak, Andrey Filev, Dimitar Tsiotras, Panagiotis School of Aerospace Engineering Institute for Robotics and Intelligent Machines Georgia Institute of Technology AtlantaGA30332-1050 United States Research & Advanced Engineering Ford Motor Company DearbornMI48121 United States

Autonomous driving has received a great deal of attention in the automotive industry and is often seen as the future of transportation. The development of autonomous driving technology has been greatly accelerated by the growth of end-to-end machine learning techniques that have been successfully used for perception, planning, and control tasks. An important aspect of autonomous driving planning is knowing how the environment evolves in the immediate future and taking appropriate actions. An autonomous driving system should effectively use the information collected from the various sensors to form an abstract representation of the world to maintain situational awareness. For this purpose, deep learning models can be used to learn compact latent representations from a stream of incoming data. However, most deep learning models are trained end-to-end and do not incorporate in the architecture any prior knowledge (e.g., from physics) of the vehicle. In this direction, many works have explored physics-infused neural network (PINN) architectures to infuse physics models during training. Inspired by this observation, we present a Kalman filter augmented recurrent neural network architecture to learn the latent representation of the traffic flow using front camera images only. We demonstrate the efficacy of the proposed model in both imitation and reinforcement learning settings using both simulated and real-world datasets. The results show that incorporating an explicit model of the vehicle (states estimated using Kalman filtering) in the end-to-end learning significantly increases performance. © 2023, CC BY-NC-SA.

关键词： Reinforcement learning

Sensitivity Analysis of Support Vector Regression-Based Incremental Capacity Analysis for Battery State of Health Estimations

学校读者我要写书评

暂无评论

Sensitivity Analysis of Support Vector Regression-Based Incr...

control Technology and Applications (CCTA),

作者： Qinan Zhou Erik Hellström Dyche Anderson Jing Sun Department of Mechanical Engineering University of Michigan Ann Arbor MI USA Research and Advanced Engineering Ford Motor Company Dearborn MI USA Department of Naval Architecture and Marine Engineering University of Michigan Ann Arbor MI USA

Lithium-ion battery state of health (SOH) estimation has become a critical challenge for battery management systems (BMS) in electrified vehicles and consumer electronics. While support vector regression (SVR)-based incremental capacity analysis (ICA) was established as an effective method for estimating SOH, the published work on this topic has been focused on a single set of charging conditions. This paper proposes an improved SVR-based ICA method that expands the working conditions of prior algorithms by incorporating the temperature and C rate sensitivities of ICA explicitly into SOH estimation models. The proposed algorithm is shown to work for different positive electrode chemistries. Moreover, range insensitivity is established for the proposed SVR-based ICA algorithm, which allows the resulting SOH estimation models to be used in various partial charging ranges encountered in practice. Less than 1% root-mean-square SOH estimation error is achieved when the proposed algorithm is assessed using publicly available datasets.

关键词：

A formal process of hierarchical functional requirements development for Set-Based Design

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Sun, Minghui Chen, Zhaoyang Bakirtzis, Georgios Jafarzadeh, Hassan Fleming, Cody Coordinated Systems Lab Department of Mechanical Engineering Iowa State University AmesIA50011 United States Autonomous Systems Group Oden Institute for Computational Engineering & Sciences The University of Texas at Austin AustinTX78712 United States Advanced Control Methods Research and Advanced Engineering Ford Motor Company DearbornMI48126 United States

The design of complex systems is typically uncertain and ambiguous at early stages. Set-Based Design is a promising approach to complex systems design as it supports alternative exploration and gradual uncertainty reduction. When designing a complex system, functional requirements decomposition is a common and effective approach to progress the design incrementally. However, the current literature on Set-Based Design lacks formal guidance in functional requirements decomposition. To bridge the gap, we propose a formal process to hierarchically decompose the functional requirements for Set-Based Design. A four-step formal process is proposed to systematically define, reason, and narrow the sets, and eventually decompose the functional requirement into the sub-requirements. Such a process can be used by the individual suppliers working in parallel at multiple levels of abstraction and guarantee that the resulting system will eventually satisfy the top-level functional requirements. An example of designing a cruise control system is applied to demonstrate the feasibility of the proposed process. © 2022, CC BY.

关键词： Large scale systems

Hierarchical Fuel-Cell Airpath control: an Efficiency-Aware MIMO control Approach Combined with a Novel Constraint-Enforcing Reference Governor

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Bacher-Chong, Eli Ayubirad, Mostafa Ali Qiu, Zeng Wang, Hao Goshtasbi, Alireza Ossareh, Hamid R. The Department of Mechanical Engineering University of Vermont BurlingtonVT05405 United States The Department of Electrical and Biomedical Engineering The University of Vermont BurlingtonVT05405 United States The Research & Advanced Engineering at Ford Motor Company DearbornMI48124 United States

This paper presents a hierarchical mutlivariable control and constraint management approach for an air supply system for a proton exchange membrane fuel cell (PEMFC) system. The control objectives are to track desired compressor mass airflow and cathode inlet pressure, maintain a minimum oxygen excess ratio (OER), and run the system at maximum net efficiency. A multi-input multi-output (MIMO) internal model controller (IMC) is designed and simulated to track flow and pressure set-points, which showed high performance despite strongly coupled plant dynamics. A new set-point map is generated to compute the most efficient cathode inlet pressure from the stack current load. To enforce OER constraints, a novel reference governor (RG) with the ability to govern multiple references (the cascade RG) and the ability to speed up as well as slow down a reference signal (the cross-section RG) is developed and tested. Compared with a single-input single-output (SISO) airflow control approach, the proposed MIMO control approach shows up to 7.36 percent lower hydrogen fuel consumption. Compared to a traditional load governor, the novel cascaded cross-section RG (CC-RG) shows up to 3.68 percent less mean absolute percent error (MAPE) on net power tracking and greatly improved worst-case OER on realistic drive-cycle simulations. control development and validations were conducted on two fuel cell system (FCS) models, a nonlinear open-source model and a proprietary ford high-fidelity model. © 2023, CC BY-NC-ND.

关键词： Efficiency

The Significance of Octane Numbers to Hybrid Electric Vehicles with Turbocharged Direct Injection Engines

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Zhou, Zhenbiao Kar, Tanmay Yang, Yi Brear, Michael Leone, Thomas Anderson, James E. Shelby, Michael Lacey, Joshua Department of Mechanical Engineering University of Melbourne ParkvilleVIC3010 Australia Ford Research and Advanced Engineering Ford Motor Company DearbornMI48121 United States Department of Mechanical Engineering KU Leuven Leuven3000 Belgium

Hybrid powertrains reduce engine operation time at light loads where the engine is less efficient. For spark-ignition engines, this means that a larger fraction of operation time could be knock-limited, and the fuel octane quality could be more important for knock control and vehicle fuel efficiency. Built on our recent work on a conventional vehicle (Zhou et al., Fuel 302, 121095), this paper extends the investigation of fuel octane and vehicle fuel efficiency to hybrid electric vehicles using a same turbocharged, direct-injection engine with the number of cylinders reduced from 4 to 3. The impact of the research Octane Number (RON) and the motor Octane Number (MON) on knock-limited fuel efficiency loss (KLFEL) is investigated using the Octane Index model, OI = (1-K)·RON + K·MON, where K is the weighting factor of the two octane numbers. Vehicle modelling is conducted to investigate the K and KLFEL distributions in three EPA drive cycles and the impact of octane numbers on these distributions. With increasing hybridization, the engine combustion is quantified by more negative K distributions and larger knock-limited fuel losses (in both relative and absolute terms). Increasing the RON and decreasing the MON are found to both improve the fuel efficiency, but the former is significantly more effective than the latter for a given change in octane number. The increasing importance of fuel octane specifications to vehicle fuel efficiency is therefore demonstrated with increasing degree of powertrain hybridization. © 2022, The Authors. All rights reserved.

关键词： Hybrid vehicles

Event-Triggered Cloud-based Nonlinear Model Predictive control with Neighboring Extremal Adaptations

学校读者我要写书评

暂无评论

Event-Triggered Cloud-based Nonlinear Model Predictive Contr...

IEEE Conference on Decision and control

作者： Amin Vahidi-Moghaddam Zhaojian Li Nan Li Kaixiang Zhang Yan Wang Department of Mechanical Engineering Michigan State University East Lansing MI USA Department of Aerospace Engineering University of Michigan Ann Arbor MI USA Research and Advanced Engineering Ford Motor Company Dearborn MI USA

ISBN: (数字)9781665467612

ISBN: (纸本)9781665467629

Model predictive control (MPC) is a popular optimal control approach that can explicitly deal with system constraints. However, its high computational cost motivates researches on various approximation methods that make the MPC more practically viable for systems with fast dynamics and/or limited on-board computing power. Neighboring extremal (NE) is one such method that can be used to adapt to a nominal open-loop MPC sequence based on state perturbations so that it alleviates the online computational burden and extends the applicability of the MPC. In this paper, a cloud-based nonlinear MPC (NMPC) framework is considered where the NMPC is first performed on the cloud by taking advantage of its powerful computation and storage capabilities. To save the computational power, instead of solving the NMPC problem at every time step, the open-loop control sequence is downloaded for on-board implementation, then the NE is employed to systematically update the control sequence based on the deviation of the measured/estimated state from the nominal state obtained by the cloud-based NMPC. However, due to unknown disturbances and request-response communication delays between the plant and the cloud, the cloud-based NMPC may not show a reasonable performance for the plant even with the NE adaptations when the deviation is large. Therefore, we develop an event-triggered scheme based on a cost criteria to strike a balance between performance and cost. Finally, the efficacy of the proposed control synthesis is demonstrated with simulations on the application of the cart inverted pendulum.

关键词： Costs Power measurement Perturbation methods Computational modeling Optimal control Time measurement Computational efficiency

Extended Neighboring Extremal Optimal control with State and Preview Perturbations

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Vahidi-Moghaddam, Amin Zhang, Kaixiang Li, Zhaojian Yin, Xunyuan Song, Ziyou Wang, Yan The Department of Mechanical Engineering Michigan State University East LansingMI48824 United States The School of Chemistry Chemical Engineering and Biotechnology Nanyang Technological University 637459 Singapore The Department of Mechanical Engineering National University of Singapore 117575 Singapore The Research and Advanced Engineering Ford Motor Company DearbornMI48121 United States

Optimal control schemes have achieved remarkable performance in numerous engineering applications. However, they typically require high computational cost, which has limited their use in real-world engineering systems with fast dynamics and/or limited computation power. To address this challenge, Neighboring Extremal (NE) has been developed as an efficient optimal adaption strategy to adapt a pre-computed nominal control solution to perturbations from the nominal trajectory. The resulting control law is a time-varying feedback gain that can be pre-computed along with the original optimal control problem, and it takes negligible online computation. However, existing NE frameworks only deal with state perturbations while in modern applications, optimal controllers (e.g., predictive controllers) frequently incorporate preview information. Therefore, a new NE framework is needed to adapt to such preview perturbations. In this work, an extended NE (ENE) framework is developed to systematically adapt the nominal control to both state and preview perturbations. We show that the derived ENE law is two time-varying feedback gains on the state perturbation and the preview perturbation. We also develop schemes to handle nominal non-optimal solutions and large perturbations to retain optimal performance and constraint satisfaction. Case study on nonlinear model predictive control is presented due to its popularity but it can be easily extended to other optimal control schemes. Promising simulation results on the cart inverted pendulum problem demonstrate the efficacy of the ENE algorithm. © 2023, CC0.

关键词： Inverted pendulum

Drivetrain clunk control via a reference governor

学校读者我要写书评

暂无评论

IFAC-PapersOnLine 2021年第20期54卷 846-851页

作者： Prithvi Reddy Mahdi Shahbakhti Maruthi Ravichandran Jeff Doering Mechanical Engineering Department Michigan Technological University Houghton USA Mechanical Engineering Department University of Alberta Edmonton CA Research and Advanced Engineering Division Ford Motor Company Dearborn USA

Presence of backlash in an automotive drivetrain leads to an undesirable phenomenon during torque delivery called clunk that adversely affects the drivability of a vehicle. Torque shaping controllers can help in reducing clunk, but may cause the vehicle to feel sluggish due to their conservative torque delivery during lash crossing. In this work, we develop a novel, model-based torque shaping clunk controller using a landing reference governor approach, which improves the drivability of a vehicle by reducing clunk while also quickly delivering the driver requested torque. The competing design requirements of the controller are framed as an optimal controls problem with hard constraints, and this optimization is performed offline. The performance of the designed controller is assessed using experimentally-validated simulation models, and the controller’s robustness is analyzed for variations in driver requested torque. Our controller is easily calibratable, and implemented as look-up tables (LUTs) for making it possible to run on embedded processors.

关键词： Automotive backlash clunk drivetrain control reference governor adaptive control optimization