Search

Summary: Recurrent neural networks (RNN) emerged as powerful tools to model and analyze the nonlinear behavior of electronic circuits accurately and quickly. Efforts to improve the accuracy of RNN will lead to the design of better‐quality products, which is essential in various fields such as the design of energy harvesting (EH) systems. EH techniques can provide the electrical energy needed for low‐power electronics without the need for a battery or with minimal dependency. Due to the importance of the active voltage balancing circuit in EH systems, we have proposed a new macromodeling method called dropout local‐feedback deep recurrent neural network (Dropout‐LFDRNN) to model and analyze this circuit along with two other nonlinear circuits as examples. This technique is an advance over the LFDRNN macromodeling method, and based on the obtained results from the measurements, we were able to build a fast macromodel for the active balancing circuit, which outperforms conventional deep recurrent neural network modeling method in terms of accuracy without sacrificing speed. It is worth mentioning that this proposed technique in this paper can be considered a viable approach for modeling and analysis of nonlinear electronic components and circuits. In addition to the advantage of generating a more accurate model, the model based on the Dropout‐LFDRNN approach is much faster than the existing transistor‐level models. [ABSTRACT FROM AUTHOR]

In order to model high-speed nonlinear circuits, recurrent neural network (RNN) has been widely used in computer-aided design (CAD) area to achieve high performance and fast models compared with the existing models. Despite their advantages, they still have challenges such as large training time and limited test accuracy. In this article, the batch normalization (BN) method is applied to deep RNN leading to a much shorter training time and more accurate models compared with the conventional RNN. The proposed BN-RNN method works by modifying the distribution of the internal nodes of a deep network in the training course as an internal auxiliary shift yielding a much faster training. Indeed, the internal covariance shift will be reduced and the training of deep neural networks will be accelerated via a normalization step applied to the layers of RNN. BN-RNN, moreover, has a beneficial effect on gradient flow through the grid by reducing the dependence of gradients on the scale of network parameters or their initial values. This provides a much better learning process without the risk of divergence. For verifying the proposed method, time-domain modeling of three high-speed nonlinear circuits operating at the GHz region is provided. Comparisons of the training and test errors between RNN and BN-RNN, and evaluation time comparisons between transistor level and the BN-RNN-based models for these circuits prove the higher speed of the models obtained from the BN-RNN method. In addition, it is shown that training using the proposed method requires much less CPU time and number of epochs. [ABSTRACT FROM AUTHOR]

Summary: In this paper, a novel method is presented for dynamic behavioral modeling of nonlinear circuits. The proposed adjoint recurrent neural network (ARNN) model is an extension of the existing recurrent neural network (RNN) technique which adds derivative information to the training data set. This addition makes training more efficient while using fewer data in comparison with the conventional RNN method with the same accuracy. Also, formulation of proposed ARNN model makes it suitable for parallel computation. Therefore, the proposed technique makes the training process much more efficient than RNN by using derivative information and parallelization. Additionally, the proposed model is much faster compared to conventional models present in existing simulation tools. The validity and accuracy of the proposed model is illustrated through macromodeling of a commercial NXP's 74LVC04A device and a five‐stage complementary metal‐oxide semiconductor (CMOS) receiver device. [ABSTRACT FROM AUTHOR]

Balancing the voltage of series connected supercapacitors is a necessity. Various passive and active balancing techniques are reported for alleviating the problems of leakage and overvoltage. In this paper, a novel active balancing approach based on boost converter is presented leading to the implementation of a piezoelectric energy harvesting (EH) system. Besides, this nonlinear boost converter is designed, implemented, and modeled using a new macromodeling approach. In this regard, data measured by the implemented boost converter passed through local feedback deep recurrent neural networks (LFDRNNs), in order to model the nonlinear behavior of this converter, and this model can be used to design the EH system. LFDRNN can be trained directly using the input–output waveform samples of the main circuit without knowing its internal details, and the obtained model has similar accuracy compared to the original circuit. The main focus of this paper is the new LFDRNN macromodeling method which is associated with the boost converter‐based active balancing technique. Our experimental results show that LFDRNN extends the ability of conventional neural network‐based models to express the dynamic behavior of nonlinear circuits while increasing the accuracy. Additionally, LFDRNN‐based models are much faster than existing models in simulation tools. [ABSTRACT FROM AUTHOR]

This article presents a new macromodeling approach for nonlinear electronic components and circuits based on long short-term memory (LSTM) neural network. LSTM proposes a more efficient training process in comparison with the conventional recurrent neural network (RNN) training. Conventional structures such as RNN suffer from the gradient vanishing problem during the training process. LSTM addresses this issue and solves this problem in an efficient way. In order to train the proposed structure, some input and output waveforms of the original circuit, called training waveforms, should be obtained from simulation tools or measurements. Model creation using proposed method does not require information on details inside the components and input–output training waveforms are sufficient to construct the model. The provided numerical results in this article show that the proposed method is more efficient than RNN techniques for modeling components and packages in terms of both speed and accuracy. The findings suggest that the proposed method significantly reduces the training time in comparison with conventional state-of-art modeling techniques. Furthermore, simulation time of the obtained model from the proposed technique is less than both conventional models (such as SPICE models) used in circuit simulation tools and models obtained from conventional RNN method. Three practical examples, namely, an audio amplifier, Texas Instruments (TIs) SN74AHCT540 device, and MOS inverter, are utilized to manifest the validity of the proposed macromodeling approach. [ABSTRACT FROM AUTHOR]

Summary: In this paper, a new method called local‐global feedback recurrent neural network (LGFRNN) is proposed for dynamic behavioral modeling of nonlinear circuits. The structure of the proposed method is based on recurrent neural network and constructed by time‐delayed local and global feedbacks. Adding time‐delayed feedbacks has a great impact on the learning capability of previous neural network‐based methods. Moreover, time‐delayed local feedbacks alleviate the problem of slow convergency of the conventional neural network‐based methods in the training phase. The proposed LGFRNN can be trained only by having sampled input‐output waveforms of the original circuit without knowing the internal details of the circuit. A training algorithm based on real‐time recurrent learning (RTRL) is used to train LGFRNN. After the training phase, the proposed LGFRNN provides accurate macromodel of a nonlinear circuit. The proposed method is more accurate compared with the conventional neural‐based models (which do not benefit from time‐delayed local‐global feedbacks) and also significantly reduces the training time of the conventional models. Moreover, proposed LGFRNN is faster than the existing models in simulation tools. The validity of the proposed method is verified by time‐domain modeling of three nonlinear devices including commercial TI's SN74AHCT540 device, five‐stage complementary metal‐oxide‐semiconductor (CMOS) receiver, and commercial TI's LM324 power amplifier. [ABSTRACT FROM AUTHOR]

Abstract This paper introduces Deep Recurrent Neural Network (DRNN) structure for time-domain modeling of nonlinear circuits. DRNN can be directly trained by sampled input-output waveforms of the original nonlinear circuit without knowing the internal details of the circuit. A training algorithm based on Real-Time Recurrent Learning (RTRL) is developed to train the proposed DRNN. After the training phase, DRNN provides accurate and fast macromodel of the nonlinear circuit. Our experimental results show that DRNN extends the ability of conventional neural-based models to express the dynamic behavior of a nonlinear circuit, and also significantly enhances the accuracy of the conventional neural circuit modeling. Moreover, DRNN is considerately faster than the existing models in simulation tools. The validity and advantages of the proposed deep learning-based method are verified by time-domain modeling of three nonlinear devices including 5-stage CMOS receiver, commercial TI's LM324 power amplifier, and TI's SN74AHCT540 device. [ABSTRACT FROM AUTHOR]