Modular Simulation Research Articles

Ecology-habitat-flow modular simulation model for the recommendation of river ecological flow combination

Ecology-habitat-flow modular simulation model for the recommendation of river ecological flow combination

Comparison of Residual-based Methods on Fault Detection

An important initial step in fault detection for complex industrial systems is gaining an understanding of their health condition. Subsequently, continuous monitoring of this health condition becomes crucial to observe its evolution, track changes over time, and isolate faults. As faults are typically rare occurrences, it is essential to perform this monitoring in an unsupervised manner. Various approaches have been proposed not only to detect faults in an unsupervised manner but also to distinguish between different potential fault types. In this study, we perform a comprehensive comparison between two residual-based approaches: autoencoders, and theinput-output models that establish a mapping between operating conditions and sensor readings. We explore the sensorwise residuals and aggregated residuals for the entire system in both methods. The performance evaluation focuses on three tasks: health indicator construction, fault detection, and health indicator interpretation. To perform the comparison, we utilize the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dynamical model, specifically a subset of the turbofan engine dataset containing three different fault types. All models are trained exclusively on healthy data. Fault detection is achieved by applying a threshold that is determined based on the healthy condition. The detection results reveal that both models are capable of detecting faults with an average delay of around 20 cycles and maintain a low false positive rate. While the fault detection performance is similar for both models, the input-output model provides better interpretability regarding potential fault types and the possible faulty components.

Modeling control strategies for prosumers in a Python-based modular simulation tool

The planned massive increase of producers and consumers such as electric vehicles, heat pumps and photovoltaic systems in distribution grids will lead to new challenges in the electrical power system. These can include grid congestions at the low voltage level but also at higher voltage levels. Control strategies can enable the efficient use of flexibilities and therefore help mitigate upcoming problems. However, they need to be evaluated carefully before their application in the energy system to avoid any unwanted effects and to choose the most fitting strategy for each application. In this publication, a Python-based modular simulation tool for developing and analysing control strategies for prosumers, which uses pandapower (Thurner et al. 2018), is presented. It is intended for sequential simulations and enables detailed operational analyses, which include evaluating the influence on grid situations, the necessary behavior of energy system components, required measurements and communications. This publication also gives an overview of control strategies, existing simulation tools, how the modular simulation tool fits in and illustrates its functionalities in an application example, which further highlights its versatility and efficiency. Time series simulations with the tool allow analyses regarding the effect of control strategies on power flow results. Moreover, the simulation tool also facilitates evaluating the behavior of energy system components (e.g. distribution substations), necessary communications and measurements as well as any faults that might occur.

An attention-based temporal convolutional network method for predicting remaining useful life of aero-engine

Researches on Remaining Useful Life (RUL) prediction of aero-engine could help to make maintenance plans, improve operation reliabilities and reduce maintenance costs. While deep learning methods have been widely used in RUL prediction research, most deep learning-based RUL prediction methods tend to treat input features as equally important. Contributions of different channels and time steps from input features are not considered simultaneously, which will inevitably affect efficiencies and accuracies of RUL prediction. Therefore, a novel deep learning-based RUL prediction method named attention-based temporal convolutional network (ATCN) is proposed in this article. First, an improved self-attention mechanism is used to weight contributions of different time steps from input features. Input features of time steps closely related to RUL are enhanced by the improved self-attention mechanism, which could improve efficiencies of feature extraction in a network. Then, a temporal convolutional network is constructed to capture long-term dependent information and extract feature representations from weighted features of the improved self-attention mechanism. Next, a squeeze-and-excitation mechanism is adopted to weight contributions of different channels from feature representations, which could help to improve prediction accuracies of the network. Finally, a fully connected layer is constructed to fuse weighted features to output RUL values. A commercial modular aero-propulsion system simulation (C-MAPSS) dataset from NASA is applied to verify effects of the proposed method. Performances of the proposed method are compared with those based on different neural network architectures, such as CNN, RNN, LSTM, DCNN, TCN, BiGRU-TSAM, AGCNN and channel attention plus Transformer. Results show that the proposed method could yield results with higher accuracy for RUL prediction of aero-engine than other methods.

MUSE-RASA captures human dimension in climate-energy-economic models via global geoAI-ML agent datasets

This article provides a combined geospatial artificial intelligence-machine learning, geoAI-ML, agent-based, data-driven, technology-rich, bottom-up approach and datasets for capturing the human dimension in climate-energy-economy models. Seven stages were required to conduct this study and build thirteen datasets to characterise and parametrise geospatial agents in 28 regions, globally. Fundamentally, the methodology starts collecting and handling data, ending with the application of the ModUlar energy system Simulation Environment (MUSE), ResidentiAl Spatially-resolved and temporal-explicit Agents (RASA) model. MUSE-RASA uses AI-ML-based geospatial big data analytics to define eight scenarios to explore long-term transition pathways towards net-zero emission targets by mid-century. The framework and datasets are key for climate-energy-economy models considering consumer behaviour and bounded rationality in more realistic decision-making processes beyond traditional approaches. This approach defines energy economic agents as heterogeneous and diverse entities that evolve in space and time, making decisions under exogenous constraints. This framework is based on the Theory of Bounded Rationality, the Theory of Real Competition, the theoretical foundations of agent-based modelling and the progress on the combination of GIS-ABM.

Aquatic Risks at the Landscape Scale: A Case Study for Pyrethroid Use in Pome Fruit Orchards in Belgium.

Procedures for environmental risk assessment for pesticides are under continuous development and subject to debate, especially at higher tier levels. Spatiotemporal dynamics of both pesticide exposure and effects at the landscape scale are largely ignored, which is a major flaw of the current risk assessment system. Furthermore, concrete guidance on risk assessment at landscape scales in the regulatory context is lacking. In this regard, we present an integrated modular simulation model system that includes spatiotemporally explicit simulation of pesticide application, fate, and effects on aquatic organisms. As a case study, the landscape model was applied to the Rummen, a river catchment in Belgium with a high density of pome fruit orchards. The application of a pyrethroid to pome fruit and the corresponding drift deposition on surface water and fate dynamics were simulated. Risk to aquatic organisms was quantified using a toxicokinetic/toxicodynamic model for individual survival at different levels of spatial aggregation, ranging from the catchment scale to individual stream segments. Although the derivation of landscape-scale risk assessment end points from model outputs is straightforward, a dialogue within the community, building on concrete examples as provided by this case study, is urgently needed in order to decide on the appropriate end points and on the definition of representative landscape scenarios for use in risk assessment.

Remaining useful life with self-attention assisted physics-informed neural network

Remaining useful life (RUL) prediction as the key technique of prognostics and health management (PHM) has been extensively investigated. The application of data-driven methods in RUL prediction has advanced greatly in recent years. However, a large number of model parameters, low prediction accuracy, and lack of interpretability of prediction results are common problems of current data-driven methods. In this paper, we propose a Physics-Informed Neural Networks (PINNs) with Self-Attention mechanism-based hybrid framework for aircraft engine RUL prognostics. Specifically, the self-attention mechanism is employed to learn the differences and interactions between features, and reasonably map high-dimensional features to low-dimensional spaces. Subsequently, PINN is utilized to regularize the end-to-end prediction network, which maps features to RUL. The RUL prediction framework termed AttnPINN has verified its superiority on the Commercial Modular AeroPropulsion System Simulation (C-MAPSS) dataset. It achieves state-of-the-art prediction performance with a small number of parameters, resulting in computation-light features. Furthermore, its prediction results are highly interpretable and can accurately predict failure modes, thereby enabling precise predictive maintenance.

An overview of the MASTODON sonar simulator

The Modular Acoustic Simulation Toolset of the Department of the Navy (MASTODON) is a feature-rich ray-based sonar simulator derived from the Personal Computer Shallow Water Acoustic Toolset (PC SWAT) codebase, developed at the Naval Surface Warfare Center Panama City Division over several decades. This paper discusses the design of this new simulation architecture and gives an overview of its features and models used. New simulation capabilities will be highlighted that enable drastic speed improvements from the addition of graphics processing unit (GPU) simulation code into MASTODON’s codebase. [Work supported by the Office of Naval Research.]

Providing a User Extensible Service-Enabled Multi-Fidelity Hybrid Cloud-Deployable SoS Test and Evaluation (T&E) Infrastructure: Application of Modeling and Simulation (M&S) as a Service (MSaaS)

Autonomous and AI-enabled systems present a challenge for integration within the System of Sys-tems (SoS) paradigm. A full system of systems (SoS) testbed is necessary to verify the integrity of a given system and preserve the modularization and accountability of its constituent systems. This integrated system needs to support iterative, continuous testing and development. This need war-rants the development of a virtual environment that provides the ground truth in a simulated sce-nario, interfaces with real-world data, and uses various domain-specific and domain-agnostic simulation systems for development, testing, and evaluation. These required features present a non-trivial challenge wherein constructive models and systems at different levels of fidelity need to interoperate to advance the testing, evaluation, and integration of complex systems. Such a virtual and constructive SoS architecture should be independent of the underlying computational infra-structure but must be cloud-enabled for wider integration of AI-enabled software components. This paper will present a modular Simulation, Experimentation, Analytics, and Test (SEAT) Lay-ered Architecture Framework, a 10-step methodology. This paper will also demonstrate a case study of a hybrid cloud-enabled simulation SoS that allows extensibility, composability, and de-ployability in different target environments.

Multiscale global and local self-attention-based network for remaining useful life prediction

Remaining useful life (RUL) prediction plays an important role in prognostics and health management (PHM) and can significantly enhance equipment reliability and safety in various engineering applications. Accurate RUL prediction enables proactive maintenance planning, helping prevent potential hazards and economic losses caused by equipment failures. Recently, while deep learning-based methods have swept the RUL prediction field, most existing methods still have difficulties in simultaneously extracting multiscale global and local degradation feature information from raw multi-sensor monitoring data. To address these issues, we propose a novel multiscale global and local self-attention-based network (MGLSN) for RUL prediction. MGLSN consists of global and local feature extraction subnetworks based on self-attention, which work in parallel to simultaneously extract the global and local degradation features of equipment and can adaptively focus on more important parts. While the global network captures long-term dependencies between time steps, the local network focuses on modeling local temporal dynamics. The design of parallel feature extraction can avoid the mutual influence of information from global and local aspects. Moreover, MGLSN adopts a multiscale feature extraction design (multiscale self-attention and convolution) to capture the global and local degradation patterns at different scales, which can be combined to better reflect the degradation trend. Experiments on the widely used Commercial Modular Aero-Propulsion System Simulation (CMAPSS), New CMAPSS (N-CMAPSS), and International Conference on Prognostics and Health Management 2008 challenge datasets provided by NASA show that MGLSN significantly outperforms state-of-the-art RUL prediction methods and has great application prospects in the field of PHM.

Integrated simulation-based calibration and sensitivity analysis of a compressed air energy storage system

Wind energy systems show tremendous potential toward the reduction of greenhouse gas (GHG) emissions; however, the rate of generation of this mode of clean energy remains predominantly intermittent, since it is produced by constantly changing natural drivers, such as wind availability and wind velocity. In this work, a novel framework is proposed which combines a modular process simulator, and a Python environment, to calibrate the operation, and perform a sensitivity analysis of a compressed air energy storage system (CAES) system. Six operational variables are identified via various Monte-Carlo simulations, and a SOBOL analysis of the results highlight three key variables that significantly influence the two primary outputs of a CAES system: the LCOE and the exergy destroyed. Our results successfully identify two novel design metrics that can inform D-CAES design and optimization, for future simulation and experimental works targeted toward wind energy capture and storage.

Modular design and simulation analysis of truss robot based on spatial bars structure

For industrial applications, truss robots can perform mutipleoperations, such as transporting objects and handling tools independently, but face the challenges of inconvenient disassembly and non-recyclability. Therefore, a truss robot load-bearing module based on a spatial bar structure is designed, which consists of truss bars and connectors. Ansys Workbench is used for the static analysis of the truss robot beam. The results show that the maximum stress of the truss robot beam is 32.46 MPa and the maximum deformation is 1.1 mm under the load of 12,000 N. The maximum stress of the beam is significantly small than the yield limit of the material and the deformation is small, which meets the requirements of structural strength and load stability of the truss robot. Based on the mode analysis of the truss robot, the inherent vibration and frequency of the truss robot are extracted, providing a reference basis for the stability control of the truss robot. The modular truss robot has mutiple advantages, including the large span, high load capacity and assembly flexibility.

Vision-Based Virtual Simulation Platform for Planetary Rovers

In order to evaluate the feasibility of planetary exploration missions, it is imperative to construct planetary terrain environments on the ground. Nevertheless, the implementation of this method is characterized by time-consuming and challenging factors. By contrast, resorting to simulation approaches represents a cost-effective and high-efficiency alternative, which can facilitate the simulation validation of planetary rover exploration missions in an effective manner. In this paper, a modular planet rover simulation platform is proposed; by building and overlaying multiple feature layers corresponding to the surface of the planet, we realize the simulation of high-resolution fine terrain and through adjusting the terrain parameters to meet the needs of different simulated terrain. The experimental results show that we have built a scene to satisfy the requirements of visual effects and physically realistic characteristics of simulation. Using an improved PatchMatch stereomethod to recover images captured real time by a planetary rover navigation camera in a virtual environment provided more complete three-dimensional terrain data for subsequent simulation validation of local path planning. Finally, a simulation environment that combines high-fidelity visual effects and kinematic characteristics supports visualizing the simulation platform: we propose a path planning method using global planning combined with local obstacle avoidance, and we obtain the optimal path that satisfies the kinematic constraints.

MCTAN: A Novel Multichannel Temporal Attention-Based Network for Industrial Health Indicator Prediction.

Health indicator prediction, such as remaining useful life prediction and product quality prediction, is an important aspect of industrial intelligence. It is essential to process the massive multichannel industrial time series collected from the Industrial Internet of Things for the industrial health indicator prediction. At present, there are still three issues that need to be considered for industrial health indicator prediction. First, it is difficult to directly connect the distant positions in the industrial time series to extract the temporal relations, which decreases the efficiency of extracting the potential long-distance temporal relations and training networks. Second, it should be fully considered that data from different channels have different contributions. Equally dealing with the contributions of each channel will weaken the representational ability of prediction networks. Third, the loss function deals with early predictions and delay predictions equally, which will lead to high risks caused by delay predictions. In this article, for these issues, a novel multichannel temporal attention-based network (MCTAN) is proposed for industrial health indicator prediction, which can weigh contributions of different channels through the channel attention while avoiding the loss of the temporal information and directly connect each time series position to the local fields of the sequence through the multi-head local attention mechanism to efficiently extract potential long-distance temporal relations. Then, a weighted mean square error loss function differently dealing with early predictions and delay predictions by setting dynamic weights is presented to reduce delay predictions. Next, to deal with the above-mentioned issues systematically, a framework combining data preprocessing and MCTAN collaboratively is introduced to predict industrial health indicators through multichannel time series. Finally, the experiments are carried out on the commercial modular aero-propulsion system simulation dataset to measure the performances, including the accuracy of industrial health indicator predictions and the inference speed.

Advancing aircraft engine RUL predictions: an interpretable integrated approach of feature engineering and aggregated feature importance

In this study, we present a comprehensive approach for predicting the remaining useful life (RUL) of aircraft engines, incorporating advanced feature engineering, dimensionality reduction, feature selection techniques, and machine learning models. The process begins with a rolling time series window, followed by the extraction of a multitude of statistical features, and the application of principal component analysis for dimensionality reduction. We utilize a variety of feature selection methods, such as Genetic Algorithm, Recursive Feature Elimination, Least Absolute Shrinkage and Selection Operator Regression, and Feature Importances from a Random Forest model. As a significant contribution, we introduce the novel aggregated feature importances with cross-validation (AFICv) technique, which ranks features based on their mean importance. We establish a selection criterion that retains features with a cumulative mean sum equal to 70%, thereby reducing the complexity of machine learning models and enhancing their generalizability. Four machine learning regression models—Natural and Extreme Gradient Boosting, Random Forest, and Multi-Layer Perceptron—were employed to evaluate the effectiveness of the selected features. The performance of our proposed method is assessed by the evaluation metrics Root Mean Square Error (RMSE) and R2 Score, and also considered within-interval percentages and relative accuracy metrics. Importantly, a novel PCA interpretability was introduced to provide real-world context and enhance the utility of our findings for domain experts. Our results indicate that the proposed AFICv technique efficiently achieves competitive performance across the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) sub-datasets using a significantly smaller subset of features, thus contributing to a more effective and interpretable RUL prediction methodology for aircraft engines.

A new multi-sensor fusion with hybrid Convolutional Neural Network with Wiener model for remaining useful life estimation

With the development of smart manufacturing, the health monitoring of the machines has become important. Remaining useful life (RUL) estimation, which could predict the future machine state, has attracted much more attentions. Deep learning (DL) based RUL has achieved remarkable results. But it still faces the issues on the multi-sensor fusion process and the health index (HI) construction, and both of these two issues can affect DL models for RUL. To overcome these issues, this research designed a new hybrid model of Convolutional Neural Network (CNN) and Wiener process, named hybrid CNN-Wiener model. First, the CNN network is adopted to achieve feature-level fusion of multi-sensor signals and to calculate the virtual HI of the machine. Second, the Wiener process model is developed to estimate the value of RUL using virtual HI. Third, the Wiener process model is designed as the layer in CNN network and trained with CNN together. The hybrid CNN-Wiener model has been tested on the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dataset, and its results show that the hybrid CNN-Wiener model has obtained the remarkable promotion by comparing with other famous DL models. The ablation studies have been tested and it shows that the hybrid CNN-Wiener model has been promotion largely with the Wiener model and the multi-sensor fusion techniques.

Fault Prognosis of Turbofan Engines

In the era of industrial big data, prognostics and health management is essential to improve the prediction of future failures to minimize inventory, maintenance, and human costs. Used for the 2021 PHM Data Challenge, the new Commercial Modular Aero-Propulsion System Simulation dataset from NASA is an open-source benchmark containing simulated turbofan engine units flown under realistic flight conditions. Deep learning approaches implemented previously for this application attempt to predict the remaining useful life of the engine units, but have not utilized labeled failure mode information, impeding practical usage and explainability. To address these limitations, a new prognostics approach is formulated with a customized loss function to simultaneously predict the current health state, the eventual failing component(s), and the remaining useful life. The proposed method incorporates principal component analysis to orthogonalize statistical time-domain features, which are inputs into supervised regressors such as random forests, extreme random forests, XGBoost, and artificial neural networks. The highest performing algorithm, ANN–Flux with PCA augmentation, achieves AUROC and AUPR scores exceeding 0.94 for each classification on average. In addition to predicting eventual failures with high accuracy, ANN–Flux achieves comparable remaining useful life RMSE for the same test split of the dataset when benchmarked against past work, with significantly less computational cost.

A predictive model of Lunar Gateway molecular outgassing and plume-induced contamination

The Gateway space station will operate both as a lunar outpost and a hub for space exploration missions in support of a long-term human presence on the Moon, and is under development by NASA in collaboration with ESA, CSA, JAXA, and international commercial partners. Gateway will experience induced molecular contamination from sources including materials outgassing, venting, and the nominal operation of its chemical and electric propulsion systems. The degradation effects of molecular contamination on spacecraft are well-documented and understood, and predictive modeling tools have been developed to aid in the mitigation of such contamination effects in the mission design phase. Therefore, the Gateway Induced Environments and Thermal teams have worked with JPL Contamination Control to develop modular simulation tools for the prediction of molecular contaminant depositions across the mission life, and for use in requirements derivation and sensitivity assessments. Free-molecular transport and deposition of outgassing products onto sensitive Gateway receiver surfaces is calculated using a viewfactor matrix approach, while International Space Station (ISS)-heritage bipropellant plume models are used to evaluate contaminant fluxes generated by the operation of thrusters on Gateway modules and the Orion spacecraft. The alteration of optical properties, e.g. the solar absorptance, of contaminated surfaces is likewise calculated using an ISS-heritage semi-empirical model premised on flight and laboratory testing data. This framework is intended to support early identification of potential integration issues and establish a baseline for incorporating improved analysis and test data as the Gateway design matures.

Dynamical modelling and experimental validation of a fast and accurate district heating thermo-hydraulic modular simulation tool

This paper presents a new thermo-hydraulic model for district heating systems simulations, which aims at being a fast and accurate tool to simulate highly renewable networks characterized by fluctuating energy profiles. The main novel aspect of the tool lies in the heat transmission modelling over long pipes based on a Lagrangian numerical approach. In comparison to other existing models, this approach significantly reduces computational time and it increases results' accuracy. The elaborated method avoids numerical diffusion in the results and consequently allows proper prediction of temperature propagation, especially in case of fast changes of fluctuating profiles. The tool is built following a modular procedural programming approach in order to facilitate the simulation of multicomponent system. Thanks to its modular structure, every components of the system is built with the same structure that is differently declined according to each component's requirements. In this way, new additional elements' models easily fit the existing ones.The model is validated under real operating conditions with hourly monitoring data of an Italian district heating network.The results show good correspondence also in the most peripheral nodes of the network, where the largest deviations are normally encountered, thus making the model a reliable and fast simulation tool for district heating network design and operational control.

Adaptively Lightweight Spatiotemporal Information-Extraction-Operator-Based DL Method for Aero-Engine RUL Prediction.

Accurate prediction of machine RUL plays a crucial role in reducing human casualties and economic losses, which is of significance. The ability to handle spatiotemporal information contributes to improving the prediction performance of machine RUL. However, most existing models for spatiotemporal information processing are not only complex in structure but also lack adaptive feature extraction capabilities. Therefore, a lightweight operator with adaptive spatiotemporal information extraction ability named Involution GRU (Inv-GRU) is proposed for aero-engine RUL prediction. Involution, the adaptive feature extraction operator, is replaced by the information connection in the gated recurrent unit to achieve adaptively spatiotemporal information extraction and reduce the parameters. Thus, Inv-GRU can well extract the degradation information of the aero-engine. Then, for the RUL prediction task, the Inv-GRU-based deep learning (DL) framework is firstly constructed, where features extracted by Inv-GRU and several human-made features are separately processed to generate health indicators (HIs) from multi-raw data of aero-engines. Finally, fully connected layers are adopted to reduce the dimension and regress RUL based on the generated HIs. By applying the Inv-GRU-based DL framework to the Commercial Modular Aero Propulsion System Simulation (C-MAPSS) datasets, successful predictions of aero-engines RUL have been achieved. Quantitative comparative experiments have demonstrated the advantage of the proposed method over other approaches in terms of both RUL prediction accuracy and computational burden.