Identification Of Complex Systems Research Articles

Koopman dynamic-oriented deep learning for invariant subspace identification and full-state prediction of complex systems

One strategy for predicting the state of nonlinear dynamical systems (typically of high dimensionality) is global linearization, such as utilizing the Koopman analysis model to transform the system state into an invariant subspace that evolves linearly. A critical challenge in the Koopman model is designing or deriving observation functions, typically nonlinear, to linearize the dynamical systems. To address the challenge, this study proposes a model called SKCAE (Skip-connected Koopman Convolutional AutoEncoder) that combines guidance of physics and data to learn observation functions accurately and efficiently. The novelties of SKCAE are twofold: (1) a coordinate-transforming main network for the construction of nonlinear observation functions, enabling the explicit identification of the low-dimensional dynamics that dominate the system's dynamical evolution, and (2) a Koopman dynamics-oriented subnetwork for quantitatively interpreting the system's intrinsic mechanisms from the frequency perspective and efficiently predicting the system state. Four numerical cases with discrete/continuous spectra on Eulerian/Lagrangrian descriptions are studied, i.e., fixed-point attractor, Duffing oscillator, fluid flow past a cylinder, and channel turbulence. Results demonstrate that SKCAE achieves significant improvement (up to a hundred times) in accuracy compared to conventional models and possesses remarkable capability in handling scenarios with data noises and/or loss, owing to its intrinsic advantage in retaining a broad range of frequency components of a dynamical system.

Chaotic attractors of discrete dynamical systems used in the core of evolutionary algorithms: state of art and perspectives

This paper delves into the intersection between discrete dynamical systems and bio-inspired metaheuristic algorithms. Chaos, an inherent phenomenon manifested by both continuous and discrete dynamic nonlinear systems has become an essential component of engineering design. This study examines the relationship between chaos and bio-inspired metaheuristic algorithms from two distinct angles: the presence of chaos within the realm of algorithms, and the application of bio-inspired algorithms for the identification, control, or synthesis of complex systems. Furthermore, this paper offers an in-depth exploration of the interplay between metaheuristics and complex nonlinear dynamics, including potential avenues for future research. The presented simulation studies, along with the design of objective functions for optimization and the implementation of metaheuristics, serve as a valuable foundation for subsequent experiments involving discrete nonlinear systems and complex systems with time delays. Such systems may benefit from multiparametric or multicriteria optimization approaches, paving the way for novel advancements in the field.

A fault location method based on ensemble complex spatio-temporal attention network for complex systems under fluctuating operating conditions

Recently, traditional fault diagnosis methods achieve fault identification by establishing a sample set covering all fault degrees of different fault components in complex systems. However, the uncertainties including environmental stresses and own physical and chemical variations lead to an infinite number of fault degrees of fault components, and the fault identification of complex systems in practical engineering applications faces the challenge of fluctuating operating conditions due to variations of rotational speed and load. To address the above problem, a fault location method based on ensemble complex spatio-temporal attention network (ECSAN) is proposed in this paper, which can identify the critical fault components of complex systems by combining an ensemble learning mechanism with excellent basic estimators. A basic estimator consists of a feature extraction module, a feature enhancement module, and a classification module. In the first module, a complex spatio-temporal backbone network with strong generalization is developed to provide spatio-temporal features containing the inherent information of the sample data for complex systems. In the feature enhancement module, a lightweight complex attention layer is constructed to enhance the effective structural information of the features and reduce the interference of their redundant information. The classification module then adopts a Softmax layer to perform fault classification. Finally, an ensemble learning mechanism is designed to integrate the basic estimators. By constructing sample weights and introducing a knowledge transfer strategy, the generalization is further improved while saving training expenses. Two datasets from different experimental platforms are concerned to verify the effectiveness and superiority of this method under various operating conditions and fault degrees. The experimental results indicate that this method achieves 99.81% accuracy on a standard dataset of the mechanical system and 98.88% accuracy on a real dataset of the closed-loop control system of the water jet propulsion device, which is superior to comparison approaches.

Complex systems modelling of UK winter wheat yield

Wheat is one of the most important global crops, understanding the drivers of wheat yield has significant societal benefits. Climate variables are particularly important in determining interannual variations in wheat yield, either as primary factors which directly influence the stages of wheat growth, or as secondary factors through their influence on pests, diseases and soil conditions. Here we present a new approach to model wheat yield; an empirical method based on nonlinear complex systems identification, known as NARMAX (Nonlinear AutoRegressive Moving Average with eXogenous inputs model). We deploy the NARMAX analytical approach for a specific site, Rothamsted, UK, where detailed meteorological variables are available, together with specific information on site conditions and crop growth stages. NARMAX yield forecasts are compared with those from the WOFOST crop model and nine state-of-the-art machine learning (ML) models; experimental results show that NARMAX outperforms all the compared methods in both prediction accuracy and model interpretability. We also develop regional wheat yield forecasts derived from a new gridded meteorological data product.The NARMAX approach produces skillful forecasts (r = 0.78) of Rothamsted wheat yield for a validation period, with small errors. The NARMAX regional forecasts, based on less specific information than WOFOST, also show a high degree of skill (r = 0.73). In addition, the predictor terms chosen for the model are identifiable and can help to give insight into potential key processes involved in the determination of wheat yield at a specific location. This approach can be extended in principle to other crop types and locations. It is straightforward and inexpensive to implement, using a limited number of meteorological predictor variables, which can be taken from site-based observations, or from gridded meteorological datasets. The method is a new tool to understand the environmental drivers of wheat yields on an annual basis.

A Strategy for Identification and Structural Characterization of Compounds from Plantago asiatica L. by Liquid Chromatography-Mass Spectrometry Combined with Ion Mobility Spectrometry.

Plantago asiatica L. (PAL) as a medicinal and edible plant is rich in chemical compounds, which makes the systematic and comprehensive characterization of its components challenging. In this study, an integrated strategy based on three-dimensional separation including AB-8 macroporous resin column chromatography, ultra-high performance liquid chromatography–quadrupole time-of-flight mass spectrometry (UHPLC-Q-TOF MS), and ultra-high performance liquid chromatography-mass spectrometry with ion-mobility spectrometry (UHPLC-IM-MS) was established and used to separate and identify the structures of compounds from PAL. The extracts of PAL were firstly separated into three parts by AB-8 macroporous resin and further separated and identified by UHPLC-Q-TOF MS and UHPLC-IM-MS, respectively. Additionally, UHPLC-IM-MS was used to identify isomers and coeluting compounds, so that the product ions appearing at the same retention time (RT)can clearly distinguish where the parent ion belongs by their different drift times. UNIFI software was used for data processing and structure identification. A total of 86 compounds, including triterpenes, iridoids, phenylethanoid glycosides, guanidine derivatives, organic acids, and fatty acids, were identified by using MS information and fragment ion information provided by UHPLC-Q-TOF MS and UHPLC-IM-MS. In particular, a pair of isoforms of plantagoside from PAL were detected and identified by UHPLC-IM-MS combined with the theoretical calculation method for the first time. In conclusion, the AB-8 macroporous resin column chromatography can separate the main compounds of PAL and enrich the trace compounds. Combining UHPLC-IM-MS and UHPLC-Q-TOF MS can obtain not only more fragments but also their unique drift times and RT, which is more conducive to the identification of complex systems, especially isomers. This proposed strategy can provide an effective method to separate and identify chemical components, and distinguish isomers in the complex system of traditional Chinese medicine (TCM).

Deep reinforcement learning for self-tuning laser source of dissipative solitons

Increasing complexity of modern laser systems, mostly originated from the nonlinear dynamics of radiation, makes control of their operation more and more challenging, calling for development of new approaches in laser engineering. Machine learning methods, providing proven tools for identification, control, and data analytics of various complex systems, have been recently applied to mode-locked fiber lasers with the special focus on three key areas: self-starting, system optimization and characterization. However, the development of the machine learning algorithms for a particular laser system, while being an interesting research problem, is a demanding task requiring arduous efforts and tuning a large number of hyper-parameters in the laboratory arrangements. It is not obvious that this learning can be smoothly transferred to systems that differ from the specific laser used for the algorithm development by design or by varying environmental parameters. Here we demonstrate that a deep reinforcement learning (DRL) approach, based on trials and errors and sequential decisions, can be successfully used for control of the generation of dissipative solitons in mode-locked fiber laser system. We have shown the capability of deep Q-learning algorithm to generalize knowledge about the laser system in order to find conditions for stable pulse generation. Region of stable generation was transformed by changing the pumping power of the laser cavity, while tunable spectral filter was used as a control tool. Deep Q-learning algorithm is suited to learn the trajectory of adjusting spectral filter parameters to stable pulsed regime relying on the state of output radiation. Our results confirm the potential of deep reinforcement learning algorithm to control a nonlinear laser system with a feed-back. We also demonstrate that fiber mode-locked laser systems generating data at high speed present a fruitful photonic test-beds for various machine learning concepts based on large datasets.

Identification of geographical origin and different parts of Wolfiporia cocos from Yunnan in China using PLS-DA and ResNet based on FT-NIR.

Wolfiporia cocos, as a kind of medicine food homologous fungus, is well-known and widely used in the world. Therefore, quality and safety have received worldwide attention, and there is a trend to identify the geographic origin of herbs with artificial intelligence technology. This research aimed to identify the geographical traceability for different parts of W. cocos. The exploratory analysis is executed by two multivariate statistical analysis methods. The two-dimensional correlation spectroscopy (2DCOS) images combined with residual convolutional neural network (ResNet) and partial least square discriminant analysis (PLS-DA) models were established to identify the different parts and regions of W. cocos. We compared and analysed 2DCOS images with different fingerprint bands including full band, 8900-6850 cm-1 , 6300-5150 cm-1 and 4450-4050 cm-1 of original spectra and the second-order derivative (SD) spectra preprocessed. From all results: the exploratory analysis results showed that t-distributed stochastic neighbour embedding was better than principal component analysis. The synchronous SD 2DCOS is more suitable for the identification and analysis of complex mixed systems for the small-band for Poria and Poriae cutis. Both models of PLS-DA and ResNet could successfully identify the geographical traceability of different parts based on different bands. The 10% external verification set of the ResNet model based on synchronous 2DCOS can be accurately identified. Therefore, the methods could be applied for the identification of geographical origins of this fungus, which may provide technical support for quality evaluation.

Set-membership parameter estimation based on Voronoi vertices

Identification of complex systems often face structural modelling errors and limited or low quality data, which hinder statistical characterisations. An alternative is the set-membership approach, where errors are assumed unknown-but-bounded. Set-membership estimation aims to find a feasible parameter set (FPS), which produces model outputs that fit within given error bounds. Most algorithms are limited to linear models, small number of parameters, or to discrete approximation of the FPS. These limitations hinder parameter estimation for relatively complex systems. We present an efficient sampling-based set-membership algorithm with low computational complexity that improves the coverage of a discrete approximation of the FPS, characterised by hyperspheres from a Voronoi diagram of the parameter space. Additionally, we suggest a measure for set-membership accuracy based on deviations between the given error bounds and the feasible model output set. Our algorithm provides a balance between accuracy and computational complexity, and a tool to investigate practical identifiability.

A new random vector functional link integrated with mayfly optimization algorithm for performance prediction of solar photovoltaic thermal collector combined with electrolytic hydrogen production system

Artificial intelligence techniques have become powerful alternatives to conventional modeling techniques in different engineering disciplines. They have been applied for modeling, control, prediction, optimization, forecasting, and identification of complex systems. In this paper, a novel optimized artificial intelligence method is developed to predict the performance of Photovoltaic/Thermal Collector (PVTC) incorporated with Electrolytic Hydrogen Production (EHP) system in terms of power output of PV, PV surface cell temperature, output temperature of cooling fluid, thermal and electrical efficiency, and hydrogen production yield. A new metaheuristic algorithm called mayfly based optimization (MO) algorithm has been implemented with Random Vector Functional Link (RVFL) network to maximize the prediction accuracy. The proposed hybrid artificial intelligence model was trained and tested using experimental data. The experiments were conducted outdoors for the proposed PVTC-EHP system operating with two different cooling fluids, namely, air and water under Indian weather conditions, and their results were compared with the predicted RVFL-MO and conventional RVFL results. Moreover, five statistical criteria were used to evaluate the performance of the investigated algorithms. The experimental results showed the hybrid PVTC-EHP system can produce a daily accumulated PV output power and hydrogen production yield of 1.66 kW/day and 3.60 kg/day for water-based PVTC-EHP system and 1.22 kW/day and 4.41 kg/day for air-based PVTC-EHP system, respectively, at a mass flow rate of 0.66 kg/min. Moreover, the statistical measures showed a perfect fit between the experimental and the proposed prediction model results. The results revealed that the root mean square error for the training phase of the RVFL and RVFL-MO was 0.25 and 0.65, respectively, while it was 1.63 and 2.04 for the testing phase, which reveals the important role of MO in determining the best parameters of RVFL network that maximize its prediction performance.

Research on Parameters Optimization of High Voltage Circuit Breaker Nozzle Based on Image Recognition and Deep Learning

Traditional modeling and optimization methods for high voltage equipment, specially, circuit breaker components and nozzles, require manual testing and improvement of a large number of parameters, and the efficiency is relatively low. The strong processing power of artificial intelligence technology in the identification and prediction of complex systems is an effective solution in such case. This paper presents the optimization approach of image recognition combined with deep learning for circuit breaker nozzle. The nozzle model was conducted using the high Mach flow model in a commercial software (COMSOL Multiphysics 5.4) to study the gas flow state and behavior during cold flow, and obtains Shock wave image recognition model of the nozzle chamber based on the convolutional neural network (CNN) method of multiscale layered features. On the basis of effective image recognition, combined with deep learning Convolutional Recurrent Neural Network (CRNN), the image sequence under different parameters is sent to the convolutional layer for feature extraction, and then the feature map is input into the loop. The prediction sequence is obtained through the layer, and finally the relationship between the kinematic parameters of the nozzle and the internal gas flow state is predicted through the prediction layer. Results indicated, according to the prediction of CRNN, the range of the throat length should be between 7 and 13 mm and the angle should be between 8 ∼ 15°. The presented method could be also used for similar materials and components, with certain universality. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

Modeling and Identification of complex systems using Neuro-Fuzzy Techniques.(Dept.E)

In this paper Adaptive-Network-Based Fuzzy Inference System (ANFIS) architecture is presented and has been used as a tool for System Identification. System Identification consists of three related steps: 1:) Structure specification, 2:) Parameter estimation. 3:) Model Validation. These steps are also discussed. We use a new method to determine the structure of the ANFIS model (Fuzzy Curves). Fuzzy Curves Help in determining the number of significant inputs from a number of candidates. We determine the number of membership functions in each input by using the subtractive clustering. Finally this method is tested on two cases.

Artificial Intelligence Applications for Friction Stir Welding: A Review

Advances in artificial intelligence (AI) techniques that can be used for different purposes have enabled it to be used in many different industrial applications. These are mainly used for modeling, identification, optimization, prediction and control of complex systems under the influence of more than one parameter in industrial applications. With the increasing accuracy of AI techniques, it has also obtained a wide application area on friction stir welding (FSW), one of the production methods developed in recent years. In this study, commonly used AI techniques for FSW, results, accuracy and superiority of AI techniques are reviewed and evaluated. In addition, an overview of AI techniques for FSW in different material combinations is provided. Considering the articles examined; It is seen that welding speed, rotational speed, the plunge depth, spindle torque, shoulder design, base material, pin design/profile, tool type are used as input parameters and tensile strength, yield strength, elongation, hardness, wear rate, welding quality, residual stress, fatigue strength are used as output parameters. As can be seen from the studies, it made important contributions in deciding what input parameters should be in order to have the output parameter at the desired value. The most common used materials for FSW are Al, Ti, Mg, Brass, Cu and so on. When FSW studies using artificial intelligence techniques were examined, it was seen that 81% of the most used materials were AL alloys and 23% of them were made with dissimilar materials. The most commonly utilized AI techniques were said to be artificial neural networks (ANN), fuzzy logic, machine learning, meta-heuristic methods and hybrid systems. As a result of the examination, ANN was the most widely used method among these methods. However, in recent years, with the exploration of new hybrid methods it was seen that hybrid systems used with ANN have higher accuracy.

Multivariable Experiment Design with Application to a Wafer Stage: a Sequential Relaxation Approach for Dealing with Element-Wise Constraints

Optimal Experiment Design (OED) is an essential aspect in accurate Frequency Response Function (FRF) identification of complex systems. The aim of this paper is to optimally design experiments for FRF identification of multivariable motion systems subject to element-wise power constraints. This design problem involves solving a non-convex and generally NP-hard optimization problem. An algorithm to solving this problem approximately is presented based on sequential semi-definite relaxations. Experimental results on a wafer stage show an improvement of the FRF quality using the proposed techniques over traditional excitation design methods.

Investigating pitching moment stall through dynamic wind tunnel test

Experimental characterisation of aircraft dynamic stall can be a challenging and complex system identification activity. In this article, the authors present a method that combines dynamic wind tunnel testing with parameter estimation techniques to study the nonlinear pitching moment dynamics of a 1/12 scale Hawk model undergoing moment stall. The instrumentation setup allows direct calculation of angular acceleration terms, such as pitch acceleration, and avoids post-processing steps involving differentiation of signals. Data collected from tests, carried out at 20 m/s and 30 m/s, are used for a brief aerodynamic analysis of the observed stall hysteresis. Then an output-error-based parameter estimation process is used to parameterise dynamic stall models and furthermore, illustrate that in a scenario where the model's heave motion is constrained. The observed nonlinear behaviour arises from the nonlinear angle of attack and linear pitch rate components.

Two-Stage Backward Elimination Method for Neural Networks Model Reduction

The single-hidden-layer neural networks (NN) has been widely used for complex system identification. However, the hidden neurons are often determined by trial-and-error method and the amount of neurons is usually large. This commonly leads to over-fitting problem and the training process is time consuming. In this paper, we propose a two-stage backward elimination (TSBE) method to obtain a parsimonious network with fewer hidden neurons but remains a good performance and saves training time. In the first stage, neural networks with a predetermined number of hidden neurons is trained based on stochastic gradient decent (SGD) algorithm with part of training data and Least absolute shrinkage and selection operator (Lasso) is applied for dropping redundant neurons leading to a simplified neural model. In the second stage, the remaining training data is used to update the parameters of the simplified neural model. A simulation example is used to validate and show that the novel approach gives a more compressed model and higher level of accuracy comparing with the recently proposed pruning-based method.

Ways of identification of complex systems

Introduction.To identify complex systems, models of different types are used depending on the goals set. The complexity of the choice of models is due to the complexity of the behavior of the systems under consideration at various times during which the basic properties of the systems can drastically change. Main part. The example of the airborne carousel E. Lorenz shows the use of fractal modelling to describe the behavior of numerically irreducible systems.It is presented an algorithm for determining the area of self-similarity of the object under consideration. According to the authors, the algorithm allows ones to reduce the probability of the object malfunctioning.Conclusions.It is considered a possibility of applying the fractal models for identification of complex systems.

The effect of multiple external representations (MERs) worksheets toward complex system reasoning achievement

The application of a systems approach to assessing biological systems provides hope for a coherent understanding of cell dynamics patterns and their relationship to plant life. This action required the reasoning about complex systems. In other sides, there were a lot of researchers who provided the proof about the instructional successions. They involved the multiple external representations which improved the biological learning. The researcher conducted an investigation using one shoot case study design which involved 30 students in proving that the MERs worksheets could affect the student's achievement of reasoning about complex system. The data had been collected based on test of reasoning about complex system and student's identification result who worked through MERs. The result showed that only partially students could achieve reasoning about system complex, but their MERs skill could support their reasoning ability of complex system. This study could bring a new hope to develop the MERs worksheet as a tool to facilitate the reasoning about complex system.

Advanced Motion Control for Precision Mechatronics: Control, Identification, and Learning of Complex Systems

Manufacturing equipment and scientific instruments, including wafer scanners, printers, microscopes, and medical imaging scanners, require accurate and fast motions. An increase in such requirements necessitates enhanced control performance. The aim of this paper is to identify several challenges for advanced motion control originating from these increasing accuracy, speed, and cost requirements. For instance, flexible mechanics must be explicitly addressed through overactuation, oversensing, inferential control, and position-dependent control. This in turn requires suitable models of appropriate complexity. One of the main advantages of such motion systems is the fact that experimenting and collecting large amounts of accurate data is inexpensive, paving the way for identifying and learning of models and controllers from experimental data. Several ongoing developments are outlined that constitute part of an overall framework for control, identification, and learning of complex motion systems. In turn, this may pave the way for new mechatronic design principles, leading to fast lightweight machines where spatio-temporal flexible mechanics are explicitly compensated through advanced motion control.

Fast Determination of Moment of Inertia of Permanent Magnet Synchronous Machine Drives for Design of Speed Loop Regulator

This paper proposes a novel method for the fast determination of moment of inertia of permanent magnet synchronous machine drive systems. It is based on the use of sinusoidal perturbation signals and can determine the combined moment of inertia within one sinusoidal cycle of perturbation while the influence of viscous friction is eliminated during the modeling process. It does not need the aid of complex system identification algorithms, and thanks to the elimination of influence of viscous friction, the proposed scheme shows higher accuracy than the conventional method without taking into account. Furthermore, its accuracy is also competitive with the conventional method using complex system identification algorithms, for example, the model reference adaptive system. Besides, the performance of designed speed regulators using the estimated mechanical parameters and the influence of mismatching of mechanical parameters are also investigated.

Modeling and Identification of Nonlinear Systems: A Review of the Multimodel Approach—Part 1

The efficacy of the multimodel framework (MMF) in modeling and identification of complex, nonlinear, and uncertain systems has been widely recognized in the literature owing to its simplicity, transparency, and mathematical tractability, allowing the use of well-known modeling analysis and control design techniques. The approach proved to be effective in addressing some of the shortcomings of other modeling techniques such as those based on a single nonlinear autoregressive network with exogenous inputs model or neural networks. A great number of researchers have contributed to this active field. Due to the significant amount of contributions and the lack of a recent survey, the review of recent developments in this field is vital. In this two-part paper, we attempt to provide a comprehensive coverage of the multimodel approach for modeling and identification of complex systems. The study contains a classification of different methods, the challenges encountered, as well as recent applications of MMF in various fields. Part 1 of this paper presents an overview of MMF for modeling and identification of nonlinear systems as well as the review of recent developments in the partitioning strategies employed.