Guarantee Robust Performance Research Articles

Active Gamma-Ray Log Pattern Localization With Distributionally Robust Reinforcement Learning

Accurately localizing 1D signal patterns, such as Gamma-ray well-log depth matching, is crucial in the oilfield service industry as it directly affects the quality of oil and gas exploration. However, traditional methods such as well-log curve analysis and pattern hand-picking matching are labor-intensive and heavily rely on human expertise, leading to inconsistent results. Although attempts have been made to automate this process, challenges such as low computational performance, non-robustness, and non-generalization remain unsolved. To address these challenges, we have developed a data-driven AI system that learns an active signal pattern localization strategy inspired by human attention. Our artificial intelligence system uses an offline reinforcement learning (RL) framework as its central component, which solves a highly abstracted Markov decision process problem via offline training on human-labeled historical data. The RL agent uses top-down reasoning to determine the location of target signal fragments by deforming a bounding window using simple transformation actions. To overcome distribution shifts between logged data and real and ensure generalization, we propose a discrete distributionally robust soft actor-critic RL framework (DRSAC-Discrete) to solve the Markov decision process problem under uncertainty. By exploring unfamiliar environments in a restrictive manner, the DRSAC-Discrete algorithm provides a safe solution that can be used when data is limited during the early stage of this industrial application. We evaluated the reinforcement learning-based localization system on augmented field Gamma-ray well-log datasets, and the results showed promising localization capability. Furthermore, the DRSAC-Discrete algorithm demonstrated relatively robust performance guarantees when facing data shortage.

Modeling and robust structural control design for hybrid AC/DC microgrids with general topology

Hybrid AC/DC microgrids are designed to utilize different types of renewable energy sources, meeting the requirements of numerous kinds of loads. A hybrid microgrid consists of a DC sub-grid, an AC sub-grid, and an interfaced sub-grid. In this paper, it is assumed that the DC and AC sub-grids contain several distributed generation (DG) units with arbitrary topology, and the interfaced region consists of several bidirectional interlink converter (IC) units. In the islanded mode of hybrid microgrid operation, one of the key control objectives is to stabilize and regulate the point of common coupling (PCC) voltages of both DC and AC sub-grids and also manage the proportional power-sharing between the sub-grids. This paper proposes a modeling framework and an optimal decentralized output-feedback-based control strategy for the voltage regulation of both DC and AC sub-grids as well as the current adjustment of IC units in a hybrid AC/DC microgrid with general topology. The proposed scheme guarantees robust stability and performance of hybrid microgrids subject to different uncertainty sources. In our proposed modeling setting, the uncertainty sources including plug-and-play (PnP) functionality of DG and IC units, microgrid topology changes, loads variations, and uncertain filter parameters of IC units are modeled as a convex polytope. The control design problem for the polytopic model is converted into a multi-objective convex optimization problem with structural constraints on decision variables. Several scenarios are performed in MATLAB/Simulink to validate the effectiveness of the proposed control mechanism for hybrid AC/DC microgrids.

Lane Compression

This article presents Lane Compression, a lightweight lossless compression technique for machine learning that is based on a detailed study of the statistical properties of machine learning data. The proposed technique profiles machine learning data gathered ahead of run-time and partitions values bit-wise into different lanes with more distinctive statistical characteristics. Then the most appropriate compression technique is chosen for each lane out of a small number of low-cost compression techniques. Lane Compression’s compute and memory requirements are very low and yet it achieves a compression rate comparable to or better than Huffman coding. We evaluate and analyse Lane Compression on a wide range of machine learning networks for both inference and re-training. We also demonstrate the profiling prior to run-time and the ability to configure the hardware based on the profiling guarantee robust performance across different models and datasets. Hardware implementations are described and the scheme’s simplicity makes it suitable for compressing both on-chip and off-chip traffic.

Supremo: Cloud-Assisted Low-Latency Super-Resolution in Mobile Devices

We present Supremo, a cloud-assisted system for low-latency image Super-Resolution (SR) in mobile devices. As SR is extremely compute-intensive, we first further optimize state-of-the-art DNN to reduce the inference latency. Furthermore, we design a mobile-cloud cooperative execution pipeline composed of specialized data compression algorithms to minimize end-to-end latency with minimal image quality degradation. Finally, we extend Supremo to video applications by formulating a dynamic optimal control algorithm to design Supremo-Opt, which aims to maximize the impact of SR while satisfying latency and resource constraints under practical network conditions. Supremo upscales 360p image to 1080p in 122 ms, which is 43.68 x faster than on-device GPU execution. Compared to cloud offloading-based solutions, Supremo reduces wireless network bandwidth consumption and end-to-end latency by 15.23 x and 4.85 x compared to baseline approach of sending and receiving whole images, and achieves 2.39 dB higher PSNR compared to using conventional JPEG to achieve similar data size compression. Furthermore, Supremo-Opt guarantees robust performance in practical scenarios.

A Theory of Experimenters: Robustness, Randomization, and Balance

This paper studies the problem of experiment design by an ambiguity-averse decision-maker who trades off subjective expected performance against robust performance guarantees. This framework accounts for real-world experimenters’ preference for randomization. It also clarifies the circumstances in which randomization is optimal: when the available sample size is large and robustness is an important concern. We apply our model to shed light on the practice of rerandomization, used to improve balance across treatment and control groups. We show that rerandomization creates a trade-off between subjective performance and robust performance guarantees. However, robust performance guarantees diminish very slowly with the number of rerandomizations. This suggests that moderate levels of rerandomization usefully expand the set of acceptable compromises between subjective performance and robustness. Targeting a fixed quantile of balance is safer than targeting an absolute balance objective. (JEL C90, D81)

Backstepping approach for design of PID controller with guaranteed performance for micro-air UAV

Flight controllers for micro-air UAVs are generally designed using proportional-integral-derivative (PID) methods, where the tuning of gains is difficult and time-consuming, and performance is not guaranteed. In this paper, we develop a rigorous method based on the sliding mode analysis and nonlinear backstepping to design a PID controller with guaranteed performance. This technique provides the structure and gains for the PID controller, such that a robust and fast response of the UAV (unmanned aerial vehicle) for trajectory tracking is achieved. First, the second-order sliding variable errors are used in a rigorous nonlinear backstepping design to obtain guaranteed performance for the nonlinear UAV dynamics. Then, using a small angle approximation and rigorous geometric manipulations, this nonlinear design is converted into a PID controller whose structure is naturally determined through the backstepping procedure. PID gains that guarantee robust UAV performance are finally computed from the sliding mode gains and from stabilizing gains for tracking error dynamics. We prove that the desired Euler angles of the inner attitude controller loop are related to the dynamics of the outer backstepping tracker loop by inverse kinematics, which provides a seamless connection with existing built-in UAV attitude controllers. We implement the proposed method on actual UAV, and experimental flight tests prove the validity of these algorithms. It is seen that our PID design procedure yields tighter UAV performance than an existing popular PID control technique.

Tustin's technique based digital decentralized load frequency control in a realistic multi power system considering wind farms and communications delays

Abstract This paper presents a digital model of decentralized Load Frequency Control (LFC) using an optimal PID controller-based Particle Swarm Optimization (PSO) in Egyptian Power System (EPS) considering communication delays. The EPS includes both conventional generation units (i.e., non-reheat, reheat, and hydraulic power plants) with inherent non-linearities and wind power, which extracted from Zafarana wind farm, location in Egypt. Thus, the optimal digital controller-based Tustin’s technique is designed for every subsystem of the EPS separately to guarantee the stability of the overall closed-loop system. The performance of the proposed digital model is tested and compared with the analog model under variation in loading patterns, loading conditions, system parameters, wind farm penetration, and communication delays. Results approved that, the proposed digital model can effectively regulate the system frequency and guarantee robust performance under different conditions. Also, it gives a reliable performance at large sampling times, which means a reduction of implementation cost.

A Novel Coordination Scheme of Virtual Inertia Control and Digital Protection for Microgrid Dynamic Security Considering High Renewable Energy Penetration

With rapid growth of renewable energy sources (RESs) in modern power systems, the microgrids (μGs) have become more susceptible to the disturbances (e.g. large frequency/voltage fluctuations) than the conventional power systems due to decreasing their inertia constant. This low system inertia issue could affect the μGs stability and resiliency in the situation of uncertainties, thus threaten their dynamic security. Hence, preserving μG dynamic security is one of the important challenges, which is addressed in this study. Therefore, this study proposes a novel concept of frequency control incorporating a virtual inertia control-based optimal proportional-integral controller to emulate virtual inertia into the μG control loop, thus stabilising μG frequency during high penetration of RESs. Moreover, the proposed virtual inertia control system is coordinated with digital over/under frequency protection for enhancement of the frequency stability and preservation of the μG dynamic security because of the high integration level of the RESs. The simulation results of the studied μG are carried out using MATLAB/Simulink ® software to validate the effectiveness of the proposed coordination scheme. Results approved that the proposed coordination scheme can effectively regulate the μG frequency and guarantee robust performance to preserve the dynamic security of μG with high penetration of RESs for different contingencies.

Development of $$ H_{\infty } $$ H ∞ -Robust Control Strategy for Drying Process in Fluidized Food Bed Dryers

Moisture control of food products is of great importance in drying processes. Improving the dryer control system results in consistent production and reduction in energy consumption. In this paper, a state observer is used to estimate the moisture content and other state variables online. Estimated states are used in a control system including the regulator for disturbance rejection and robust controller for tracking objective. Since the bed dryer mechanism is a multivariable complex system associated with various uncertainties, a realistic uncertain model is considered. Moisture content and temperature of tea leaves (or other food products) are controlled at desired set points by manipulating heating rate and food feed rate. Modeling the uncertainties and developing a code based on \( \mu \)-synthesis with DK-iteration, an optimal robust controller is designed. It is observed that controller guarantees robust stability and performance of the system in tracking of desired set points, in the presence of model uncertainties.

Modelling and robust attitude trajectory tracking control of 3-DOF four rotor hover vehicle

PurposeThe purpose of this paper is to present μ-synthesis-based robust attitude trajectory tracking control of three degree-of-freedom four rotor hover vehicle.Design/methodology/approachComprehensive modelling of hover vehicle is presented, followed by development of uncertainty model. A μ-synthesis-based controller is designed using the DK iteration method that not only handles structured and unstructured uncertainties effectively but also guarantees robust performance. The performance of the proposed controller is evaluated through simulations, and the controller is also implemented on experimental platform. Simulation and experimental results validate that μ-synthesis-based robust controller is found effective in: solving robust attitude trajectory tracking problem of multirotor vehicle systems, handling parameter variations and dealing with external disturbances.FindingsPerformance analysis of the proposed controller guarantees robust stability and also ensures robust trajectory tracking performance for nominal system and for 15-20 per cent variations in the system parameters. In addition, the results also ensure robust handling of wind gusts disturbances.Originality/valueThis research addresses the robust performance of hover vehicle’s attitude control subjected to uncertainties and external disturbances using μ-synthesis-based controller. This is the only method so far that guarantees robust stability and performance simultaneously.

Robust coordinated transmission for cooperative small cell networks

Within a macrocell with a large coverage area, multiple small cells are deployed such that each small cell base station (SBS) supports wireless service demands from user equipments (UEs). Each UE can be simultaneously served by multiple SBSs for quality of service (QoS) enhancement. When there exist hotspot areas with a number of UEs, the SBSs near the hotspot areas may experience a higher resource utilisation level than those outside of the hotspot areas, resulting in a shortage of available resources. The authors propose a robust resource-utilisation-based coordinated transmission for heterogeneous networks with a locally different level of traffic demands. In the utilisation-based coordinated transmission, low-utilisation SBSs with a small number of UEs are selected to serve a newly joining UE because they have more capacity to serve requests with bursty traffic demand. They further formulate the selection of cooperative SBSs as a robust optimisation problem in order to ensure that UEs have sufficiently high signal-to-interference-plus-noise ratios, even with channel estimation inaccuracy and strong interference from non-cooperative SBSs. The simulation results indicate that the proposed method guarantees robust and efficient service performance in heterogeneous small cell networks.

Robust load‐frequency control in interconnected power systems

This study describes a robust load-frequency control (LFC) scheme in a two-area interconnected power system with uncertain disturbances. A sector-bounded H∞ control approach is used for compensating of the non-linearities based on a minimum order dynamic LFC model to avoid the complexity of a high-order model. The simulation in time domain from the use of the proposed controller has been compared with the use of a conventional proportional–integral (PI) type controller affected by the same load change. The simulated results have proved the significant improvement over the PI controller and have verified the effectiveness of the proposed robust H∞ controller which can minimise the frequency deviations significantly. This approach treats the tie-line power exchanges in the presences of external disturbance, parameter uncertainties and accidental crashing of generating units as the structural or model uncertainties discussed in this study. Simulation results also show that the proposed H∞ controller guarantees robust performance and fast response under a wide range of operating conditions and system uncertainties.

Generic Nonsmooth <inline-formula> <tex-math notation="LaTeX">$\mathcal {H}_{\infty }$ </tex-math></inline-formula> Output Synthesis: Application to a Coal-Fired Boiler/Turbine Unit With Actuator Dead Zone

Actuator nonidealities, such as backlash, dead zone, and others, present in a number of industrial systems, are known to severely degrade system performance. Providing nonconservative closed-loop robust performance guarantees for these systems in a consistent manner has been an open problem. For example, in boiler/turbine units, the turbine valve position actuation for manipulating steam flow rate is characterized by a small mismatch between the turbine valve command and the actual valve position, producing a small steady-state regulation error in the plant outputs. The standard linear $\mathcal {H}_{\infty }$ controller designed to provide zero steady-state error regulation drives this error to zero, producing the undesirable oscillations in the control signals and the plant outputs. This paper develops a nonsmooth $\mathcal {H}_{\infty }$ output regulator theory addressing this problem and applies this theory to the experimentally validated boiler/turbine model with actuator dead zone. The simulation results showing a considerable performance improvement are given.

Suppression of nonlinear regenerative chatter in milling process via robust optimal control

During the milling process, self-excited vibration or chatter adversely affects tool life, surface quality and productivity rate. In this paper, nonlinear cutting forces of milling process are considered as a function of chip thickness with a complete third order polynomial (instead of the common linear dependency). An optimal control strategy is developed for chatter suppression of the system described through nonlinear delay differential equations. Counterbalance forces exerted by actuators in x and y directions are the control inputs. For optimal control problem, an appropriate performance index is defined such that the regenerative chatter is suppressed while control efforts are minimized. Optimal control law is determined based on variation of extremals algorithm. Results show that under unstable machining conditions, regenerative chatter is suppressed effectively after applying the optimal control strategy. In addition, optimal controller guarantees robust performance of the process in the presence of model parametric uncertainties.

FCS/nORB: A feedback control real-time scheduling service for embedded ORB middleware

Object Request Broker (ORB) middleware has shown promise in meeting the functional and real-time performance requirements of distributed real-time and embedded (DRE) systems. However, existing real-time ORB middleware standards such as RT-CORBA do not adequately address the challenges of (1) managing unpredictable workload, and (2) providing robust performance guarantees portably across different platforms. To overcome this limitation, we have developed software called FCS/nORB that integrates a Feedback Control real-time Scheduling (FCS) service with the nORB small-footprint real-time ORB designed for networked embedded systems. FCS/nORB features feedback control loops that provide real-time performance guarantees by automatically adjusting the rate of remote method invocations transparently to an application. FCS/nORB thus enables real-time applications to be truly portable in terms of real-time performance as well as functionality, without the need for hand tuning. This paper presents the design, implementation, and empirical evaluation of FCS/nORB. Our extensive experiments on a Linux testbed demonstrate that FCS/nORB can provide deadline miss ratio and utilization guarantees in the face of changes in platform and task execution times, while introducing only a small amount of overhead.

H∞ Control for Singularly Perturbed Bilinear Systems with Parameter Uncertainties Using Successive Galerkin Approximation

This paper presents a new algorithm for the closed-loop H∞ composite control of singularly perturbed bilinear systems with time-varying parameter uncertainties and exogenous disturbance using the successive Galerkin approximation (SGA). The singularly perturbed bilinear system is decomposed into two subsystems of a slow-time scale and a fast-time scale via singular perturbation theory, and then two H∞ control laws are obtained for each subsystem. H∞ control theory guarantees robust closed-loop performance but the resulting problem is difficult to solve for bilinear systems. In order to overcome the difficulties inherent in the H∞ control problem, the suitable robust H∞ feedback control law can be constructed in term of the approximated solution to a Hamilton-Jacobi-Isaac equation using SGA. The composite control law consists of H∞ control laws for each subsystem.

3203 多様性を有する場に対応する新たなロバスト設計法の構築(OS11 創発性と多様性の設計)

The need for robust design that enables the products to ensure robust performance in diverse conditions has been focused. Robust design guarantees robust performance in manufacturing variation of products and diverse use environment. In this study, new robust design method which is applicable to diverse design problems was proposed. In this method, the robust index, the weighted robust index, and the adjusted robust index are used as evaluation indexes of robustness. And confirming the effectiveness, the proposed robust design method was applied to a public seat design. As a result of the verification, the solution obtained by the proposed robust design method was better than the existing method, and it was confirmed that the general versatility of the proposed method.

Genetic-Algorithm-Based Fixed-Structure Robust H∞Loop-Shaping Control of a Pneumatic Servosystem

The robust controller designed by conventional H∞optimal control is complicated, high-order, and difficult to implement practically. In industrial applications, structures such as lead-lag compensators and PID are widely used because their structure is simple, tuning parameters are fewer, and they are lower-order. Their disadvantages are that control parameters are difficult to tune for good performance and they lack robustness. To solve these problems, we propose an algorithma genetic-algorithm-based fixed-structure robust H∞loop-shaping controlfor designing the robust controller. Conventional H∞loop shaping is a sensible procedure for designing the robust controller. To obtain parameters in the proposed controller, we proposed a genetic algorithm to optimize specified-structure H∞loop shaping problem. The infinity norm of transfer function from disturbances to states is minimized via searching and evolutionary computation. The resulting optimal parameters stabilize the system and guarantee robust performance. We applied the evolutionary robust controller to a pneumatic servosystem. To compare performance, we studied three types of controller PID with a derivative first-order filter controller, a PI controller, and an H∞loop-shaping controller. Results of experiments demonstrate the advantages of a simple structure and robustness against parameters changing. Simulations verify the effectiveness of the proposed technique.

On the Performance of Space–Time Codes Over Spatially Correlated Rayleigh Fading Channels

We derive the exact pairwise error probability (PEP) for space-time coding over quasi-static Rayleigh fading channels in the presence of spatial fading correlation. We show that receive correlation always degrades the PEP for all signal-to-noise ratios (SNRs). We quantify the effect of receive correlation by employing the notion of "majorization". We show that the stronger the receive correlation, the worse the PEP for all SNRs. We show that at low SNR, depending on the codes, transmit correlation can either improve or degrade the PEP performance. We show that to guarantee robust performance for arbitrary transmit correlation, the minimum eigenvalue of the codeword pair difference matrix should be maximized among all codeword pairs.

Robust Control of Identified Models with Mixed Parametric and Non-Parametric Uncertainties

A framework for identification oriented robust controller design is developed. The model is identified from open-loop i/o-data and contains parametric uncertainties as well as an additive and norm-bounded error. The set of all robustly stabilising controllers, that additionally guarantee robust performance is characterised by a system of second order cones, which can be efficiently solved by interior point algorithms.