Black-box Modeling Method Research Articles

A Study on the Frequency-Domain Black-Box Modeling Method for the Nonlinear Behavioral Level Conduction Immunity of Integrated Circuits Based on X-Parameter Theory.

During circuit conduction immunity simulation assessments, the existing black-box modeling methods for chips generally involve the use of time-domain-based modeling methods or ICIM-CI binary decision models, which can provide approximate immunity assessments but require a high number of tests to be performed when carrying out broadband immunity assessments, as well as having a long modeling time and demonstrating poor reproducibility and insufficient accuracy in capturing the complex electromagnetic response in the frequency domain. To address these issues, in this paper, we propose a novel frequency-domain broadband model (Sensi-Freq-Model) of IC conduction susceptibility that accurately quantifies the conduction immunity of components in the frequency domain and builds a model of the IC based on the quantized data. The method provides high fitting accuracy in the frequency domain, which significantly improves the accuracy of circuit broadband design. The generated model retains as much information within the frequency-domain broadband as possible and reduces the need to rebuild the model under changing electromagnetic environments, thereby enhancing the portability and repeatability of the model. The ability to reduce the modeling time of the chip greatly improves modeling efficiency and circuit design. The results of this study show that the "Sensi-Freq-Model" reduces the broadband modeling time by about 90% compared to the traditional ICIM-CI method and improves the normalized mean square error (NMSE) by 18.5 dB.

Research on Black-Box Modeling Prediction of USV Maneuvering Based on SSA-WLS-SVM

Unmanned surface vessels (USVs) are required to perform motion prediction during a task. This is essential for USVs, especially when conducting motion control, and this work has been proven to be complicated. In this paper, an off-line black box modeling method for USV maneuvering, the Sparrow search algorithm-based weighted-least-squares support vector machine (SSA-WLS-SVM) was proposed to recognize the motion model of a USV. Firstly, the construction of the USV test platform and the processing process of the experimental data were introduced, the correctness of the MMG model was verified using a comparison of the test data and the simulation results, and then the MMG model was used to produce sample data later. To improve the stability and robustness of LS-SVM, weighted least squares and SSA were introduced to perform the optimization of the parameters of the algorithm and its kernel function. Then, the random maneuvering dataset was obtained using simulation on the MMG model, which was then preprocessed and used for training the black-box model. To verify the generalization ability of the identified model, the black-box model was used for comparison analysis between motion prediction with the proposed model and maneuvering test on the USV platform in a scenario different from the training data.

Conducted Coupling Model and Anti-Interference Filter Design for Digital Control Circuit Under Electrical Fast Transient Noise

Control and protection equipment working in complex electromagnetic environment often suffers from electrical fast transient (EFT) noise, which easily leads to the failure of digital control circuits of the equipment. In order to realize low-cost and effective anti-interference solutions to the failure problem, it is necessary to study the anti-interference quantitative design method of the digital control circuit. In this article, a conducted interference coupling model of the digital control circuit under EFT noise is established. As an important part of the coupling path, the impedance model of the chip is built. A black-box impedance modeling method of voltage regulator chip is proposed based on online measurement. The equivalent impedance model of the digital signal processing (DSP) chip is represented by the high-frequency model of the inductor. The coupling model is validated by comparing the measured results and simulated results of the terminal load port voltage. Then, a measurement and analysis method is presented to obtain the circuit failure threshold. By analyzing EFT test results at different frequencies and amplitudes, the influencing factor of the circuit failure is clarified, and the circuit failure threshold is obtained. Finally, based on the coupling model and the circuit failure threshold, a quantitative design method of an anti-interference filter is developed. The method is validated through filter prototyping for a 5 kHz/+2.0 kV EFT test.

Identification of Surrogate Models for the Prediction of Degrees of Freedom within a Tolerance Chain

The computation of assembly tolerance information is necessary to fulfill robust design requirements. This assembly is computationally costly, with current calculations taking several hours. We aim to identify surrogate models for predicting degrees of freedom within a tolerance chain based on point connections between assembly components. Thus, replacing part of the current computation workflow and consequently reduce computation time. We use manufacturing tolerances set by norms and industrial standards to identifly these surrogate models, which define all relevant features and resulting output variables. We use black-box modeling methods (artificial neural networks and gradient boosted trees), as well as white-box modeling (symbolic regression by genetic programming). We see that these three models can reliably predict the degrees of freedom of a tolerance chain with high accuracy (R2 > 0.99).

Electrical machines surrogate‐based design optimization based on novel waveform targeting strategy with improvement of the computational efficiency

Electrical machine design optimization is an expensive procedure as it contains numerous variables and multiple objectives. Therefore, it might require hundreds of time-consuming finite element analyses (FEA). Surrogate models are one of the superior alternatives that can overcome the computational burden of FEA. However, in optimization problems with sensitive input-output relationships, surrogate models can lack accuracy or suffer from unreasonable initial FE-computational cost. To tackle this problem, this work presents a novel strategy called Waveform-Targeted Surrogate Modeling (WTSM) that improves the computational efficiency of surrogate-based design optimization of electrical machines by modifying the model construction process. In this paper, an Interior Permanent Magnet Machine (IPMSM) is studied to evaluate the proposed strategy's performance compared to the conventional Black Box Modeling (BBM) method. A two-step assessment procedure consisting of multiple scenarios has been developed to analyse the reliability of the WTSM concerning different training-validation datasets. Meanwhile, the Kriging model and Latin-Hypercube-Sampling (LHS) method are employed for surrogate model construction purposes. From the discussion and the results, it can be found that the proposed WTSM strategy can significantly increase the accuracy of the surrogate modelling procedure while the required computational cost can be reduced.

Black-Box Mathematical Modeling and Heuristic Optimizing of Hampson-Linde Cycle Based on Joule-Thomson Effect and Ohm’s Law for Thermal Circuits

Nowadays, due to recent global climate changes, replacing non-environmental friendly technology with more sustainable energy is desired. Thus, researchers are designing new, environmentally friendly products, for which some use the black box modeling method. Therefore, the presented work represents a Hampson-Linde cryogenic cooler model based on Joule-Thomson Effect and Ohm’s Law for thermal circuits, optimized using a parallel “Particle Swarm Optimization” (PSO) algorithm. An innovative feature of this model is that it uses two translations—from electrical to the thermal domain and a simplifying time domain—and is implemented to provide results using less demanding computational resources and simulation time. Furthermore, a possibility for superconductive states is presented for commonly used category II superconductors. Last, but not least, based on the predicted output temperature, models of more complex processes could be developed, such as a model for the Hyperloop concept.

Dynamic Equivalence Modeling for Microgrid Cluster by Using Physical-Data-Driven Method

In practical application of microgrid cluster, the lack of full detailed information cause the failure of dynamic modeling. Although some data-driven black-box modeling method can tackle this problem, insufficient usage of prior known physical information may reduce the modeling accuracy. To tackle this challenge, a hybrid physical-data-driven method is proposed for the dynamic behavior modeling of microgrid cluster. Motivated by the equivalence of recurrent neural network (RNN) and differential equations, the differential-algebraic equations (DAEs) of unknown part are represented by gate recurrent unit (GRU) based neural network. The DAEs of prior known physical stage are embedded into the proposed neural network, which avoid unnecessary model training of prior known section and improving the modeling efficiency. At first, the basic idea of RNN based dynamic modeling is explained. Then, the modeling guidelines including data preparation and parameter design are suggested. Finally, the effectiveness of the proposed method is confirmed by a test system formed by three microgrids under grid fault and operating point changing conditions.

Residential microgrid optimization using grey-box and black-box modeling methods

The use of data-driven black-box HVAC and water heater models in the optimization of a single-home residential microgrid is investigated. Data-driven models do not rely on the specification of system geometry or thermal parameters, instead using time-variant, system-specific data to develop models of each controllable appliance. A comparison of modeling accuracy of grey-box models based on those found in the literature and black-box models developed using extreme gradient boosting is made. A multi-objective optimization problem is then developed with the objectives of minimizing energy cost and consumption while maximizing thermal comfort and the consumption of locally generated PV energy. Results demonstrate that data-driven black-box models have the potential to be used in place of grey-box models, which are more prone to the introduction of error by the user. Optimization using the grey-box and black-box modeling methods was performed using a genetic algorithm to verify the similarity of results using the two modeling methods. Optimization resulted in a maximum total energy savings during the optimized period of 32.3%, a maximum total peak energy reduction of 39.8%, and a maximum cost reduction of 38.1%.

Modeling the end-use performance of alternative fuels in light-duty vehicles

Present study investigates the end-use performance of alternative liquid fuels in the current fleet of unmodified light-duty vehicle (LDV) engines. Two mathematical models have been developed that represent the way that various fuel properties affect fuel consumption in spark-ignition (SI) and compression-ignition (CI) engines. Fuel consumption is represented by the results from the New European Driving Cycles (NEDC) in order to reflect the end-use impact. Data-driven black-box modeling and multilinear regression methods were applied to obtain both models. Additionally, quantitative analysis was performed to ensure the statistical significance of inputs (p-value below 5%). Fuel consumption (output) of various alternative fuels can be estimated with high accuracy (coefficient of determination above 0.96), knowing fuel properties (inputs) such as lower heating value, density, cetane/octane number, and oxygen content. The validation procedures confirmed the quality of predictions for both models with the average error being below 2.3%. The model performance for the examined fuels such as hydrotreated vegetable oil (HVO) and ethanol blends showed significant CO2 reduction with high accuracy. Moreover, both models could be used to estimate CO2 tailpipe emissions and are applicable to various liquid SI/CI fuels for LDV engines.

Black-Box Behavioral Modeling of Voltage and Frequency Response Characteristic for Islanded Microgrid

The voltage and frequency response model of microgrid is significant for its application in the design of secondary voltage frequency controller and system stability analysis. However, most models developed for this aspect are complex in structure due to the difficult mechanism modeling process and are only suitable for offline identification. To solve these problems, this paper proposes a black-box modeling method to identify the voltage and frequency response model of microgrid online. Firstly, the microgrid system is set as a two-input, two-output black-box system and can be modeled only by data sampled at the input and output ports. Therefore, the simplicity of modeling steps can be guaranteed. Meanwhile, the recursive damped least squares method is used to realize the online model identification of the microgrid system, so that the model parameters can be adjusted with the change of the microgrid operating structure, which makes the model more adaptable. The paper analyzes the black-box modeling process of the microgrid system in detail, and the microgrid platform, including 100 kW rated power inverters, is employed to validate the analysis and experimental results.

Accurate modeling of photovoltaic modules using a 1-D deep residual network based on I-V characteristics

Accurate and reliable modeling of photovoltaic (PV) modules is significant for optimal design, operation and evaluation of PV systems. PV models can be classified into equivalent circuit-based white box models and data-driven black box models. Due to the difficulty to obtain the ground true model parameters and the limitation posed by the predetermined model structure, white-box modeling methods generally suffer relatively low accuracy and generalization performance for arbitrary operating conditions. In addition, reported black-box models are based on the conventional artificial neural networks (ANN) that are efficient but have limited performance. In this study, motivated by the high performance of fast developing deep learning techniques, we propose a novel black-box modeling method for the PV modules using a new modified one-dimensional deep residual network (1-D ResNet) and measured I-V characteristic curves, which can predict a whole I-V curve at a time for arbitrary operating conditions. To alleviate the overfitting issue caused by imbalanced data, original I-V curve datasets with highly non-uniform operating conditions are resampled by a grid sampling approach to obtain the datasets with relatively uniform conditions for the subsequent modeling. The proposed 1-D ResNet based model is comprehensively verified and compared with a proposed single-diode based white-box model and three other conventional ANN based black-box models, using large datasets of measured I-V characteristic curves from the National Renewable Energy Laboratory (NREL). Experimental results indicate that black-box models are generally better than the white-box model. Especially, the proposed 1-D ResNet based PV model is obviously superior to other three conventional ANN based black-box models, in terms of accuracy, generalization performance and reliability.

Pan-Tilt Modelling for Face Detection

Object detection is one of the important application in the surveillance system. In system design, a model is needed to be used for the compensator design process. The black-box modeling method is used so that we can get the mathematical model of the plant without using physics laws, while it is used an experimental approach so that it takes input and output data from the open-loop system response. In modeling the system, two datasets are required to estimate the model and validate the data. In the process, Arduino Mega controller commanded to transmit Pulse Width Modulation (PWM) signals on the pan and tilt servo motors. The command sets motion to pan and tilt systems which installed a camera in it, and the dataset retrieval process can be performed. In this case, the PWM signal as input and the center coordinates of the face (pixels) as outputs based on the viola jones method for the image processing of the open-loop system. The results of the mathematical model of the system show that the data obtained is quite accurate, it is proved from the estimation and data validation that has been done.

Adaptive Black-Box Equivalent Modeling of Multiple Marine Micro-Grids Interconnected Through Subsea Cables

The interconnection of different kinds of marine micro-grids (MMGs) with plug-and-play functionality is becoming popular, and a combined model is necessary to achieve the control functionalities and operation of the system constitute of several MMGs. However, due to the remoteness of these MMGs, it is difficult to collect necessary data to establish a whole model for MMGs. Besides, subsea cables used to connect MMGs add additional challenges to the modeling because their parameters will change with ageing. Therefore, an adaptive black-box equivalent modeling method for multiple MMGs interconnected through subsea cables without knowing detailed information is proposed in this paper. And the parameters of the equivalent model of the subsea cables and the MMGs are estimated by a statistical algorithm based on the measurement data at the point of common coupling. Furthermore, case studies demonstrate the feasibility and effectiveness of the proposed methodology.

Fault detection in the distillation column process using Kullback Leibler divergence

Chemical plants are complex large-scale systems which need designing robust fault detection schemes to ensure high product quality, reliability and safety under different operating conditions. The present paper is concerned with a feasibility study of the application of the black-box modeling method and Kullback Leibler divergence (KLD) to the fault detection in a distillation column process. A Nonlinear Auto-Regressive Moving Average with eXogenous input (NARMAX) polynomial model is firstly developed to estimate the nonlinear behavior of the plant. Furthermore, the KLD is applied to detect abnormal modes. The proposed FD method is implemented and validated experimentally using realistic faults of a distillation plant of laboratory scale. The experimental results clearly demonstrate the fact that proposed method is effective and gives early alarm to operators.

Black-box modeling of residential HVAC system and comparison of gray-box and black-box modeling methods

In this article, black-box models of the residential heating, ventilation and air conditioning (HVAC) system are developed. The data of the input and output of the system is measured and the models of the energy recovery ventilator (ERV), air handling unit (AHU), buffer tank (BT), radiant floor heating (RFH) and ground source heat pump (GSHP) are developed using the system identification techniques in Matlab®. The developed models include models based on multiple-input and multiple-output (MIMO) artificial neural network (ANN), transfer function (TF), process, state-space (SS) and autoregressive exogenous (ARX) ones of each HVAC subsystem (ERV, AHU, BT and RFH). The gray-box models of the same HVAC subsystems were developed in [1] which are also compared with the black-box models developed in this article. The models were compared visually and analytically. Ranks of the models were calculated based on their relative performance. It was found that the ANN outperforms the other modeling methods followed by the ARX, TF, SS, process and gray-box models respectively.

Black-Box Behavioral Modeling and Identification of DC–DC Converters With Input Current Control for Fuel Cell Power Conditioning

Fuel cells are a prime candidate for alternative power source of future on-board power distribution systems such as those for the more-electric-aircraft and electric vehicles. These systems are comprised of a large number of power converters, often provided by a variety of manufacturers. Modeling and simulation are powerful tools to evaluate the system-level behavior but, due to confidentiality of manufacturers, black-box behavioral models of the converters are required instead of the conventional ones. This paper proposes a black-box modeling and identification method of dc–dc converters with input current control for fuel-cell power conditioning. The model is simple, requires low computational cost for simulation, and the identification procedure is based on simple experiments. Moreover, the model does not represent the internal structure of the converter, so it can be provided by the manufacturer while protecting confidential data. The method is illustrated in detail and validated by making use of a commercial dc–dc converter specifically designed for fuel-cell power conditioning.

Modeling of Gan Hemt by Using an Improved K-Nearest Neighbors Algorithm

A novel black-box modeling method for field effect transistor (FET) based on an improved K-Nearest Neighbors algorithm is proposed in this paper. K-Nearest Neighbors algorithm, which has simple algorithm structure and high accuracy, has shown great potential in regression application. A Taylor series expansion method is employed in nonlinear elements modeling of FET to improve the physical meaning of black-box model. A GaN HEMT power device is used to demonstrate the proposed method. And the experimentation shows that the calculated results using improved K-Nearest Neighbors algorithm based model fit the measurement results well.