Search

In recent years, a new paradigm in control engineering, called networked control system (NCS), has emerged due to the success of internet and associated communication infrastructure. The NCS offers various advantages compared to traditional control systems which include increase of agility, wiring reduction, and easy maintenance and diagnosis. With an objective to investigate the effectiveness of various controllers in networked environment, the present study designs a -based single-bit PID controller for the rotor side converter (RSC) of a doubly-fed induction generator (DFIG) based wind energy conversion system (WECS). This consists of 20-DFIGs of 5MW capacity each. A comparative performance investigation has been carried out with three other controllers; a conventional PID, Q-Learning and Dynamic Fuzzy Q-Learning (DFQL) controller. The study considers various imperfections of a Zigbee protocol based practical communication network such as random delay and packet dropout. The results of both the simulation and experiment demonstrate that the -based single-bit PID controller shows better performance compared to other controllers.

This paper proposes a Delta-Modulator ( $$\Delta $$ -M)-based quantised state feedback controller for Takagi–Sugeno (T–S) fuzzy networked systems. The $$\Delta $$ -M (a single bit quantiser) essentially belongs to one class of sliding mode quantisers (SMQs) and offers various advantages which include lower design complexity, lower cost and less noisy. For a prescribed quantisation error, the gains of the state-feedback controller and the quantiser are derived (both in continuous and discrete time domains) using linear matrix inequalities (LMIs) which ensures the stability of the overall system. The performance of the quantised control system is illustrated considering a practical communication network based on ZigBee protocol. The results of the simulation demonstrate that the proposed $$\Delta $$ -M-based quantised controller could effectively achieve desired performance under various imperfections of the practical communication network.

Manufacturing of composites using automated fiber placement (AFP) is a complex process that involves large number of processing conditions and variables. Improper selection of these parameters adversely affects the quality and integrity of the manufactured laminates. Thus, it is important to develop a predictive model that can assess how changes in critical process conditions alter the outputs of the manufacturing process. The goal of this investigation is to learn the complex behavior of composites by developing an intelligent model, which can subsequently be used for the prediction of various characteristics of the composites. However, manufacturing of AFP composites is both expensive and time-consuming and therefore the available data samples are less, from the prospective of machine learning, which leads to the small data learning problem . This article first solves this problem through virtual sample generation, and then a neural network based predictive model is developed to accurately learn the complex relationships between various processing parameters in AFP.

Renewable energy sources generated locally are becoming increasingly popular in order to achieve carbon neutrality in the near future. Some of these sources are being used in neighbourhood (local, or energy communities) grids to achieve high levels of self-sufficiency. However, the objectives of the local grid and the distribution grid to which it is connected are different and can sometimes conflict with each other. Although the distribution grid allows access to all variable resources, in certain circumstances, such as when its infrastructure is overloaded, redispatch measures need to be implemented. The complexity and uncertainties associated with current and future energy systems make this a challenging bi-level multi-criteria optimisation problem, with the distribution grid representing the upper level and the neighbourhood grid representing the lower level. Solving these problems numerically is not an easy task. However, there are new opportunities to solve these problems with less computational costs if we decompose the flexibility in the lower lever. Therefore, this paper presents a mathematical approach to optimise grid management systems by aggregating flexibility from neighbourhood grids. This mathematical approach can be implemented with centralised or decentralised algorithms to solve congestion problems in distribution grids.

The present study proposes a new dynamic two-level quantizer, called as hybrid delta modulator ( $\Delta _{H} $-M), which combines the features of both delta-modulator and delta-sigma modulator. In the transient state, the $\Delta _{H} $-M exhibits the dynamical behaviour of delta modulator (Δ-M) while in steady state, its behaviour is similar to delta-sigma modulator ( $\Delta \Sigma $-M). This study investigates about various dynamics of the proposed $\Delta _{H} $-M in both continuous and discrete-time domains. The stability conditions of $\Delta _{H} $-M are derived using the theory of sliding and quasi-sliding mode. The theoretical results are validated through extensive simulations. Abbreviations: Δ-M: delta modulator; $\Delta _{H} $-M: hybrid delta modulator; $\Delta \Sigma $-M: delta sigma modulator; ADC: analogue to digital converter; CT: continuous-time; DAC: digital to analogue converter; DT: discrete-time; NCS: networked control system; QSM: quasi-sliding model; QSMD: quasi-sliding mode domain; SS: steady state; TP: transient process

An adaptive neuro-fuzzy inference system (ANFIS) controller is developed and presented in this study to control hybrid energy storage system (HESS) which combines the battery and super-capacitor (SC). The battery compensates the energy requirement for a longer duration while the SC limits the stress on battery caused by the power fluctuations during transient period which alternately gives longer life span for the battery while regulate the DC link voltage constant. The proposed ANFIS controller is being compared for performance with various other controllers including the reinforcement controller based on Q-learning proportional and integral (PI) controller, fuzzy controller and conventional PI controller. Further, the state of charge (SOC) of the battery and SC are monitored in order to decide the required optimal amount of power or energy for the HESS in deficit/excess modes. The results of the simulation, in different loading conditions, indicate that the ANFIS's controller performance for the DC microgrid is superior compared to others.

In control systems applications, controllers for different plants are usually designed with different methods. Although plants may share common characteristics, these controllers are generally designed in isolation. The problem of continually learning a sequence of related tasks has been extensively studied recently. A challenge in continual learning is the phenomenon of catastrophic forgetting of knowledge of previous tasks which have been integrated into a neural network model. In this paper we evaluate the feasibility of modelling different controllers using continual learning. We explore regression versions of state-of-the-art methods and demonstrate that even the simplest continual learning approach decreases the overall Mean Average Error (MAE) by \(39\%\) of the MAE achieved by a non-continual strategy. Furthermore, a method based on dynamically expanding the network can achieve an overall MAE which is only \(18\%\) of the non-continual MAE. We also propose a set of new metrics that allow us to characterise the nature of catastrophic forgetting experienced while using different continual learning methods.

During the past two decades, a new paradigm in control engineering, called networked control system (NCS), has emerged. This has been shown to be a successful alternative to traditional control systems. The present study designs a reinforcement learning controller based on the concept of Dynamic Fuzzy Q-Learning (DFQL). This controls the rotor side converter (RSC) of a doubly-fed induction generator (DFIG) of a wind energy conversion system (WECS) through a communication network. A comparative performance investigation has been carried out with two other controllers; namely a conventional PID and Q-Learning controller. The performance of these controllers is studied considering various imperfections of the communication network such as random delay and packet dropout. Simulations are carried out considering a WECS consisting of 20-DFIGs of 5MW capacity each. The results of the study demonstrate that the DFQL controller shows better performance compared to Q-Learning and conventional PID controllers.

Manufacturing of thermoplastic composites using automated fibre placement (AFP) machine with specific characteristics is a challenging task due to the interdependence of various processing conditions and variables. It is of interest to know the accurate value of different input variables which would give the desired characteristics (outputs) of the laminates. This problem comes under the framework of inverse identification and is often ill-posed and its solution becomes increasingly difficult when the available data samples are very less. The present study develops a neural network-based inverse predictive model for AFP based manufacturing process using virtual sample generation (VSG) techniques. The efficacy of the developed predictive inverse model has been established considering varieties of experimental data. The proposed approach can be applied to a large class of manufacturing processes to determine the input conditions to a get product with desired characteristics.

The present study designs a novel ΔΣ-based PID controller for controlling the rotor side converter (RSC) of a doubly-fed induction generator (DFIG) based wind energy conversion system (WECS). The proposed controller essentially uses ΔΣ modulators to represent the analogue signals in single-bit which is ideally suited for FPGA or application specific integrated circuit implementation and offers numerous advantages due to its inherent parallel nature. The performance of the controller is compared with various controllers including a conventional PID, Q and Dynamic Fuzzy-Q learning (DFQL) based intelligent controllers. The results of the simulation on WECS consists of 20 wind turbines and 20 DFIGs, illustrate that the ΔΣ based PID controller could perform better compared to conventional PID and Q-learning based controllers.

During the past decade, networked control systems (NCS) has emerged as a viable alternative to traditional control systems due to various advantages it offers which include a reduction in system wiring, increase of system agility etc. However, the performance of various existing controllers such as PID degrades in the networked environment due to the existence of random time-varying delay, packet-dropouts which may cause instability. The present study designs a neural network (NN) based controller for NCS and investigates its performance under random time-varying delay, packet-dropouts. The performance of this controller is compared with both the classical PID and ΔΣ-based PID controllers. The robustness of the NN based controllers in the networked environment is studied under different degree of parametric uncertainties considering an example of a DC servo mechanism. The results of the comparative investigation demonstrate that the performance of the NN based controller is superior compared to other controllers.