In this paper, we present a method to analyze local and global stability in offset-free setpoint tracking using neural network controllers and we provide ellipsoidal inner approximations of the corresponding region of attraction. We consider a feedback interconnection of a linear plant in connection with a neural network controller and an integrator, which allows for offset-free tracking of a desired piecewise constant reference that enters the controller as an external input. Exploiting the fact that activation functions used in neural networks are slope-restricted, we derive linear matrix inequalities to verify stability using Lyapunov theory. After stating a global stability result, we present less conservative local stability conditions (i) for a given reference and (ii) for any reference from a certain set. The latter result even enables guaranteed tracking under setpoint changes using a reference governor which can lead to a significant increase of the region of attraction. Finally, we demonstrate the applicability of our analysis by verifying stability and offset-free tracking of a neural network controller that was trained to stabilize a linearized inverted pendulum.
Active Building Research Programme（主动建筑研究计划）；Imperial College London（帝国理工学院）
In this work, we develop a novel data-driven model predictive controller using advanced techniques in the field of machine learning. The objective is to regulate control signals to adjust the desired internal room setpoint temperature, affected indirectly by the external weather states. The methodology involves developing a time-series machine learning model with either a Long Short Term Memory model (LSTM) or a Gradient Boosting Algorithm (XGboost), capable of forecasting this weather states for any desired time horizon and concurrently optimising the control signals to the desired set point.…
While many theoretical works concerning Adaptive Dynamic Programming (ADP) have been proposed, application results are scarce. Therefore, we design an ADP-based optimal trajectory tracking controller and apply it to a large-scale ball-on-plate system. Our proposed method incorporates an approximated reference trajectory instead of using setpoint tracking and allows to automatically compensate for constant offset terms. Due to the off-policy characteristics of the algorithm, the method requires only a small amount of measured data to train the controller. Our experimental results show that this…
We propose model-plant mismatch learning offset-free model predictive control (MPC), which learns and applies the intrinsic model-plant mismatch, to effectively exploit the advantages of model-based and data-driven control strategies and overcome the limitations of each approach. In this study, the model-plant mismatch map on steady-state manifold in the controlled variable space is approximated via a general regression neural network from the steady-state data for each setpoint. Though the learned model-plant mismatch map can provide the information at the equilibrium point (i.e., setpoint), …
Optimal Volt/VAR control (VVC) in distribution networks relies on an effective coordination between the conventional utility-owned mechanical devices and the smart residential photovoltaic (PV) inverters. Typically, a central controller carries out a periodic optimization and sends setpoints to the local controller of each device. However, instead of tracking centrally dispatched setpoints, smart PV inverters can cooperate on a much faster timescale to reach optimality within a PV inverter group. To accommodate such PV inverter groups in the VVC architecture, this paper proposes a bi-level opt…
Reliable and secure operation of power systems becomes increasingly challenging as the share of volatile generation rises, leading to largely changing dynamics. Typically, the architecture and structure of controllers in power systems, such as voltage controllers of power generators, are fixed during the design and buildup of the network. As replacing existing controllers is often undesired and challenging, setpoint adjustments, as well as tuning of the controller parameters, are possibilities to counteract large disturbances and changing dynamics. We present an approach for fast and computati…
This paper deals with the design of nonlinear MPC controllers that provide offset-free setpoint tracking for models described by Neural Nonlinear AutoRegressive eXogenous (NNARX) networks. The NNARX model is identified from input-output data collected from the plant, and can be given a state-space representation with known measurable states made by past input and output variables, so that a state observer is not required. In the training phase, the Incremental Input-to-State Stability (δISS) property can be forced when consistent with the behavior of the plant. The δISS property is then levera…
Institute for Dynamic Systems and Control, ETH Zürich（苏黎世联邦理工学院动态系统与控制研究所）
This paper presents an offset-free model predictive controller for fast and accurate control of a spherical soft robotic arm. In this control scheme, a linear model is combined with an online disturbance estimation technique to systematically compensate model deviations. Dynamic effects such as material relaxation resulting from the use of soft materials can be addressed to achieve offset-free tracking. The tracking error can be reduced by 35% when compared to a standard model predictive controller without a disturbance compensation scheme. The improved tracking performance enables the realiza…
The requirement for continual improvement of idle speed control (ISC) performance is increasing due to the stringent regulation on emission and fuel economy these days. In this regard, a low-complexity offset-free explicit model predictive control (EMPC) with constraint horizon is designed to regulate the idle speed under unmeasured disturbance in presence of system delay with rigorous formulation. Particularly, we developed a high-fidelity 4-stroke gasoline-direct injected spark-ignited engine model based on first-principles and test vehicle driving data, and designed a model predictive ISC s…
Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；ETH Zürich（苏黎世联邦理工学院）；University of Stuttgart（斯图加特大学）
Fast feedback control and safety guarantees are essential in modern robotics. We present an approach that achieves both by combining novel robust model predictive control (MPC) with function approximation via (deep) neural networks (NNs). The result is a new approach for complex tasks with nonlinear, uncertain, and constrained dynamics as are common in robotics. Specifically, we leverage recent results in MPC research to propose a new robust setpoint tracking MPC algorithm, which achieves reliable and safe tracking of a dynamic setpoint while guaranteeing stability and constraint satisfaction.…
Recurrent neural network-based robust control systems with regional properties and application to MPC design
Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano（电子、信息与生物工程系，米兰理工大学）；Istituto di Sistemi e Tecnologie Industriali Intelligenti per il Manifatturiero Avanzato, Consiglio Nazionale delle Ricerche（先进智能制造工业系统与技术研究所，国家研究院）
The general context of this work is the feedback control of an infinite-dimensional system so that the closed-loop system satisfies a fading-memory property and achieves the setpoint tracking of a given reference signal. More specifically, this paper is concerned with the Proportional Integral (PI) regulation control of the left Neumann trace of a one-dimensional reaction-diffusion equation with a delayed right Dirichlet boundary control. In this setting, the studied reaction-diffusion equation might be either open-loop stable or unstable. The proposed control strategy goes as follows. First, …
The introduction of unexpected system disturbances and new system dynamics does not allow guaranteed continuous system stability. In this research we present a novel approach for detecting early failure indicators of non-linear highly chaotic system and accordingly predict the best parameter calibrations to offset such instability using deep machine learning regression model. The approach proposed continuously monitors the system and controller signals. The Re-calibration of the system and controller parameters is triggered according to a set of conditions designed to maintain system stability…
PID control has been the dominant control strategy in the process industry due to its simplicity in design and effectiveness in controlling a wide range of processes. However, traditional methods on PID tuning often require extensive domain knowledge and field experience. To address the issue, this work proposes an automatic PID tuning framework based on reinforcement learning (RL), particularly the deterministic policy gradient (DPG) method. Different from existing studies on using RL for PID tuning, in this work, we consider the closed-loop stability throughout the RL-based tuning process. I…
