DSpace community: 電機工程學系暨研究所

Electromagnetic Imaging for Buried Conductors Using Deep Convolutional Neural Networks

Tue, 23 Jun 2026 04:05:47 GMT

title: Electromagnetic Imaging for Buried Conductors Using Deep Convolutional Neural Networks abstract: In the past, many conventional algorithms, such as self-adaptive dynamic differential evolution and asynchronous particle swarm optimization, were used to reconstruct buried objects in the frequency domain; these were unfortunately time-consuming during the iterative, repeated computing process of the scattered field. Consequently, we propose an innovative deep convolutional neural network approach to solve the electromagnetic inverse scattering problem for buried conductors in this paper. Different shapes of conductors are buried in one half-space and the electromagnetic wave from the other half-space is incident. The shape of the conductor can be reconstructed promptly by inputting the received scattered fields measured from the upper half-space into the deep convolutional neural network module, which avoids the computational complexity of Green’s function for training. Numerical results show that the root mean square error for differently shaped—circular, elliptical, arrow, peanut, four-petal, and three-petal—reconstructed images are, respectively, 2.95%, 3.11%, 17.81%, 15.10%, 14.14%, and 15.24%. Briefly speaking, not only can circular and elliptical buried conductors be reconstructed; some irregular shapes can be reconstructed well. On the contrary, the reconstruction result by U-Net for buried objects is worse since it is not able to obtain a good preliminary image by processing only the upper scattered field—that is, rather than the full space. In other words, our proposed deep convolutional neural network can efficiently solve the electromagnetic inverse scattering problem of buried conductors and provide a novel method for the microwave imaging of the buried conductors. This is the first successful attempt at using deep convolutional neural networks for buried conductors in the frequency domain, which may be useful for practical applications in various fields such as the medical, military, or industrial fields, including magnetic resonance imaging, mine detection and clearance, non-destructive testing, gas or wire pipeline detection, etc.

Microwave Imaging for Half-Space Conductors Using the Whale Optimization Algorithm and the Spotted Hyena Optimizer

Tue, 23 Jun 2026 04:05:44 GMT

title: Microwave Imaging for Half-Space Conductors Using the Whale Optimization Algorithm and the Spotted Hyena Optimizer abstract: This research implements the whale optimization algorithm (WOA) and spotted hyena optimizer (SHO) in inverse scattering to regenerate the conductor shape concealed in the half-space. TM waves are irradiated from the other half-space to a perfect conductor with an unknown shape buried in one half-space. The scattered field measured outside the conductor surface with the boundary condition is used to reconstruct the object using the WOA and SHO algorithms. Several scenarios of reconstruction accuracy were compared for the WOA and SHO. The numerical simulations prove that the WOA has a better reconstruction capability.

Comparison of U-Net and OASRN Neural Network for Microwave Imaging

Tue, 23 Jun 2026 04:05:42 GMT

title: Comparison of U-Net and OASRN Neural Network for Microwave Imaging abstract: U-Net and Object-Attentional Super-Resolution Network (OASRN) neural network for electromagnetic imaging are compared and investigated in this paper. The outcome shows that though under limited training data, the regeneration capability is still highly reliable. We first transmit the electromagnetic waves to the scatterer and use the received scattered field information to calculate the estimated permittivity distribution by Green’s function, subspace method and Dominant Current Scheme (DCS). The estimation technique can effectively reduce the training process of the neural network modules. Next, we train the U-Net and OASRN modules for real-time images. Lastly, we used Root Mean Square Error (RMSE) and Structural Similarity Index Measure (SSIM) to compare and analyze the reconstructed images of the two neural networks. Numerical results show that the reconstructed image by OASRN is better than that by U-net with 5% or 20% Gaussian noise for different dielectric constant distributions.

Electromagnetic imaging of Uniaxial objects by Artificial Intelligence Technology

Tue, 23 Jun 2026 04:05:40 GMT

title: Electromagnetic imaging of Uniaxial objects by Artificial Intelligence Technology abstract: The electromagnetic (EM) imaging of uniaxial objects by the artificial intelligence (AI) technology is presented in this article. We study the 2-D inverse scattering problem from uniaxial objects illuminated by the transverse magnetic (TM) and transverse electric (TE) polarized incident waves. As the uniaxial objects have different components of permittivity along different transverse directions, the problem of TE polarization will be more severe than that of TM polarization. We use the dominant current scheme (DCS) and backpropagation scheme (BPS) to calculate the preliminary permittivity distribution. By combining with deep learning and neural networks, the permittivity distribution of those uniaxial objects can be reconstructed more accurately. U-Net is used to reconstruct the permittivity distribution because U-Net has shared the weights and biases, which can effectively reduce the network complexity and is very suitable for solving image processing problems. In the numerical results, we added different noises to compare the reconstruction results of the DCS and BPS initial estimations through the U-Net. Numerical results show that the reconstruction permittivity for the DCS initial estimation is better than that for the BPS initial estimation. Our diversity is that we have reconstructed the uniaxial objects by neural network successfully with less time-consuming effort and real-time imaging.

Wi-Fi 6E Antenna Design for All Metal Housing of Notebook

Tue, 23 Jun 2026 04:05:38 GMT

title: Wi-Fi 6E Antenna Design for All Metal Housing of Notebook abstract: In this paper, we present two antenna structures with the Wi-Fi 6E band, namely, dual-slot and single-slot antennas. The presented antennas can be applied for the all-metal housing on notebook computers and also meet the requirements of notebook computer antenna design for industry. First, we introduce the design of the dual-slot antenna at the top of the metal case. The size of the dual-slot antenna was 53 × 6 × 0.6 mm3. To meet the specifications of commercially available 13-inch laptops, we chose a 305 × 205 × 1 mm3 metal case in the simulation environment. To make our proposed antenna design meet the requirement that the reflection coefficient of the Wi-Fi 6E frequency band is lower than -10 dB, we adjusted the grounded parasitic element and antenna structure to achieve the coupling effect of the slot. This antenna achieved a high screen-to-body ratio and narrow bezels and conformed with current notebook design trends. Next, because some laptops have IR cameras that require a smaller antenna, we introduce the single-slot antenna for use above the metal case. The size of the simulated metal case was also 305 × 205 × 1 mm3, and a monopole antenna with dimensions of 30 × 4.5 × 0.6 mm3 was used. By bending the geometry of the slot antenna, it can be modified to change the coupling effect. The antenna not only achieved a high screen-to-body ratio and narrow bezels, but was also smaller than the dual-slot antenna. The two proposed antenna architectures have the advantage of compactness, with the need to open only one or two slots on the metal case of a notebook computer for the Wi-Fi 6E band.

Beamforming Relay for Millimeter-wave SWIPT System

Tue, 23 Jun 2026 04:05:36 GMT

title: Beamforming Relay for Millimeter-wave SWIPT System abstract: This research aims to develop an ultra-wideband millimeter-wave system with Simultaneous Wireless Information and Power Transfer (SWIPT), and Wireless Power Transmission (WPT) for the relays. As a large attenuation of a millimeter-wave path requires beamforming technology and relay for transmission, high- and low-resolution phase adjustments are used for optimizing beamforming. And a multi-objective function with the preset highest Bit Error Rate (BER) for SWIPT is presented. We discovered that the optimization of the beamforming by applying the adaptive differential algorithm has increased the harvesting power. To do so, we optimize the radiation pattern to meet the BER constraint for SWIPT and increase the harvested power ratio for the system simultaneously. In other words, our algorithms focus on increasing the harvesting power as soon as the information criteria is achieved. With WPT and SWIPT, the ratio of the total energy harvesting for the high-resolution array antennas is two times larger than that for the low-resolution ones. Numerical results also show that the harvesting power for the relay pointing to multiple targeted antennas simultaneously is about two times larger than that of pointing to each antenna by the time division techniques.

Optimization for Indoor 6G Simultaneous Wireless Information and Power Transfer System

Tue, 23 Jun 2026 04:05:34 GMT

title: Optimization for Indoor 6G Simultaneous Wireless Information and Power Transfer System abstract: Antenna beamforming for Simultaneous Wireless Information and Power Transfer (SWIPT) and Wireless Power Transfer (WPT) in an indoor 6G communication system is presented in this paper. The objective function is to maximize the total harvesting power for the SWIPT and WPT nodes with the constraints of the bit error rate and minimum harvesting power. In the study, the power-splitting ratio between harvesting power and decoding information can be adjusted for the SWIPT node. Due to the non-convex problem, we use Self-Adaptive Dynamic Differential Evolution (SADDE) to optimize the designed multi-objective function. We use a symmetric antenna array to study three situations of distance—closer, farther, and similar—between the transmitting antenna and the individual SWIPT and WPT nodes in this paper. Experimental results show that the overall harvesting efficiency is improved, especially in the case of SWIPT nodes closer to the transmitter. The total harvesting power can be improved by 86.7% in the total short-distance case, and by 7.87% in the total long-distance case.

Different Object Functions for SWIPT Optimization by SADDE and APSO

Tue, 23 Jun 2026 04:05:31 GMT

title: Different Object Functions for SWIPT Optimization by SADDE and APSO abstract: Multiple objective function with beamforming techniques by algorithms have been studied for the Simultaneous Wireless Information and Power Transfer (SWIPT) technology at millimeter wave. Using the feed length to adjust the phase for different objects of SWIPT with Bit Error Rate (BER) and Harvesting Power (HP) are investigated in the broadband communication. Symmetrical antenna array is useful for omni bearing beamforming adjustment with multiple receivers. Self-Adaptive Dynamic Differential Evolution (SADDE) and Asynchronous Particle Swarm Optimization (APSO) are used to optimize the feed length of the antenna array. Two different object functions are proposed in the paper. The first one is the weighting factor multiplying the constraint BER and HP plus HP. The second one is the constraint BER multiplying HP. Simulations show that the first object function is capable of optimizing the total harvesting power under the BER constraint and APSO can quickly converges quicker than SADDE. However, the weighting for the final object function requires a pretest in advance, whereas the second object function does not need to set the weighting case by case and the searching is more efficient than the first one. From the numerical results, the proposed criterion can achieve the SWIPT requirement. Thus, we can use the novel proposed criterion (the second criterion) to optimize the SWIPT problem without testing the weighting case by case.

Phase Optimization Scheme of RIS-assisted MC-OTFS Wireless Communication Systems

Tue, 26 May 2026 04:05:15 GMT

title: Phase Optimization Scheme of RIS-assisted MC-OTFS Wireless Communication Systems

Edge-Aware, Data-Efficient Fine-Tuning of Progressive GANs for Multiband Antennas

Thu, 21 May 2026 04:05:10 GMT

title: Edge-Aware, Data-Efficient Fine-Tuning of Progressive GANs for Multiband Antennas abstract: This study proposes a data-efficient fine-tuning strategy for multi-band antenna synthesis using a Wasserstein Auxiliary-Guided Progressive Growing GAN (WAG-PGGAN). Starting from a pretrained 512 × 512 dual-band PIFA-like generator trained on 4180 samples at 2.45/5.2 GHz, we introduce three 3.5-GHz wideband seeds augmented to 836 images (new:legacy ≈ 1:5) and fine-tune only the highest-resolution stage on the combined 5016-image corpus. A Hough-transform-based edge-enhancement module with an edge-aware loss preserves conductor boundaries and strengthens frequency–geometry correlation. Across n = 8 fabricated prototypes, all achieve |S11| < −10 dB and collectively span 1.86–5.83 GHz; measured total efficiencies are 52–87% (e.g., 73.6% @ 2.68 GHz, 66.7% @ 3.56 GHz, 69.0% @ 5.83 GHz), with radiation patterns consistent with simulation. The method retains prior 2.45/5.2 GHz performance while adding 3.5-GHz wideband behavior using ≤ 17% new data (836/5016), demonstrating effective transfer from small datasets. On an RTX 3060 Ti, inference is ≈ 3 s/design after ~192 h of training. Simulation–measurement agreement confirms that fine-tuned WAG-PGGAN yields high-resolution, physically valid multi-band antennas with reduced data and computational cost.

應用PGGAN和增強特徵映射的機器學習於雙頻天線設計

Thu, 21 May 2026 04:05:08 GMT

title: 應用PGGAN和增強特徵映射的機器學習於雙頻天線設計 abstract: This paper presents a systematic antenna design methodology that integrates machine learning, leveraging the progressive growth technique of Progressive Growing of GANs (PGGAN) to generate images of various dual-band PIFA-like antenna structures. The process involves using data augmentation methods to generate 4180 antenna samples. In the latent space, the authors employ Latin Hypercube Sampling to maintain diversity and combine it with the Hough Transform to enhance the edge features of the antennas while providing labelling functionality. This labelling method strengthens the relationship between antenna frequency and wavelength characteristics. The paper clearly outlines the design process, starting from the simulation of two types of single-frequency PIFA-like antennas (2.45 and 5.2 GHz, respectively) to the completion of PGGAN's generation task, resulting in a novel dual-band Wi-Fi PIFA-like antenna structure. The measurement results of the dual-band antennas exhibit excellent consistency with the simulation results.

Energy-dependent carrier masses of zinc blende semiconductors: Solved using the Kane approximation

Fri, 15 May 2026 04:05:21 GMT

title: Energy-dependent carrier masses of zinc blende semiconductors: Solved using the Kane approximation abstract: The mathematical expressions for the effective carrier masses of zinc blende semiconductors are derived, with particular emphasis on their dependence on carrier energy. In this work, Hamiltonian matrix diagonalization is employed to obtain the band energies, followed by the Kane approximation to derive analytical expressions for energy-dependent effective masses. Unlike conventional approaches that assume constant effective mass near the band edge, the present formulation explicitly accounts for band non-parabolicity. InAs and GaAs are used as representative examples to verify consistency between the derived expressions and calculated band structures. The results provide a physically transparent and analytically tractable framework for modeling carrier transport in high-speed electronic and optoelectronic devices

Mechanism Design and Control of Humanoid Robot

Fri, 27 Mar 2026 06:17:50 GMT

title: Mechanism Design and Control of Humanoid Robot

Bipedal Robot Locomotion Using Deep Reinforcement Learning

Fri, 27 Mar 2026 06:17:46 GMT

title: Bipedal Robot Locomotion Using Deep Reinforcement Learning

小型人形機器人之上階梯動作規劃

Wed, 25 Mar 2026 04:08:45 GMT

title: 小型人形機器人之上階梯動作規劃

A metasurface-integrated ultra-wideband antenna design for broadband gain stability

Thu, 19 Mar 2026 04:05:30 GMT

title: A metasurface-integrated ultra-wideband antenna design for broadband gain stability abstract: This paper presents a novel electromagnetic metasurface-integrated antenna design that utilizes periodic structures to significantly enhance gain stability across the entire ultra-wideband (UWB) frequency range (3.1–10.6 GHz). Unlike conventional approaches that position metasurfaces as separate external elements, the proposed design directly integrates the metasurface into the antenna structure. This integration not only reduces spatial requirements but also improves overall compactness, making the antenna more practical for real-world implementation. The proposed antenna demonstrates consistent gain performance and a highly uniform broadside gain response throughout the full UWB spectrum. These advantages effectively address the persistent issue of gain fluctuation found in conventional UWB antennas, which is critical for maintaining stable communication performance. Furthermore, this work introduces a systematic design methodology that balances wideband impedance matching with enhanced radiation characteristics, advancing the application of metasurface technologies in antenna engineering. The proposed technique shows promising potential for future wireless communication systems and precision indoor positioning applications.

Lightweight design and precise tracking control for differential drive wheeled mobile robot

Wed, 11 Mar 2026 04:05:23 GMT

title: Lightweight design and precise tracking control for differential drive wheeled mobile robot abstract: Differential-drive wheeled mobile robot (DDWMR) offers excellent maneuverability, making it increasingly valuable to researchers in recent years, but the control poses major challenges due to the nonlinearity nature of the system. This study proposes an integrated scheme consisting of a backstepping controller and an adaptive sliding-mode controller (ASMC) to address these challenges including the DDWMR being subject to external disturbances and uncertainties. The presented DDWMR design is lightweight, structurally robust, and mechanically simple with a total wight of about 10 kg. The new control scheme can achieve accurate trajectory tracking while ensuring dynamic stability of the DDWMR. Performance comparison of trajectory tracking was made on the approach with that of proportional-integral-derivative (PID) controller and sliding mode controller (SMC). Compared to a SMC + backstepping controller, simulation results revealed that performance improvements in trajectory tracking reached root-mean-square errors (RMSE) reduction by 3.7% in the X-axis, and by 23.1% in the Y-axis, respectively. Finally, experimental results verified that the proposed controller outperformed both traditional controllers by achieving a reduction of RMSE by 12.1% in the X-axis and by 19.1% in the Y-axis, respectively in comparison with the SMC + backstepping controller in a real-world application.

Cyber-physical System-based Wide-area IoT for Illegal Forest Logging Monitoring and Alert System

Tue, 10 Mar 2026 04:11:10 GMT

title: Cyber-physical System-based Wide-area IoT for Illegal Forest Logging Monitoring and Alert System

Deep Learning-Enhanced Iterative Modified Contrast Source Method for Electromagnetic Imaging in Half-Space

Tue, 10 Mar 2026 04:11:00 GMT

title: Deep Learning-Enhanced Iterative Modified Contrast Source Method for Electromagnetic Imaging in Half-Space abstract: This paper presents a hybrid inversion framework that integrates a physics-informed iterative algorithm with a deep learning-based refinement strategy to address the electromagnetic inverse scattering problem of a uniaxial object buried in lossy half-space environments. Specifically, an Iterative Modified Contrast Scheme (IMCS) is developed to accelerate convergence and produce stable initial estimates, yielding improved performance compared to conventional contrast source methods. These estimates are subsequently refined by U-Net architecture, thereby enhancing the image quality of the reconstructed dielectric targets. Numerical simulations demonstrate that the proposed framework achieves robust and high-fidelity reconstructions of buried high-contrast dielectric objects, even in the presence of 20% additive Gaussian noise.

DESIGN OF ACTIVE HEAT DISSIPATION SYSTEM FOR ADAPTIVE WAVELET NEURAL NETWORK CONTROL

Fri, 06 Mar 2026 04:06:07 GMT

title: DESIGN OF ACTIVE HEAT DISSIPATION SYSTEM FOR ADAPTIVE WAVELET NEURAL NETWORK CONTROL abstract: This paper develops an Adaptive Wavelet Neural Network Control (AWNNC) algorithm for radar active heat dissipation system. The radar core processor belongs to a highly precision component which consists of the electronic device of radio frequency integrated circuit (RFIC) with high power and high performance. The radar core processor should be operated in a narrowly closed environment without convection, which will increase the heat sink effect inside the core processor and further affect its reliability and life-time. The AWNNC comprises a wavelet neural network (WNN) controller and a robust compensator. The WNN controller is a principal tracking controller which is utilized to mimic an ideal controller; and the parameters of WNN are online tuned by the derived adaptation laws based on the gradient descent method. The robust compensator is designed to dispel the approximation error between the ideal controller and the WNN controller, thus the asymptotic stability of the closed–loop system can be achieved. Based on National Instruments-PCI extensions for instrumentation (NI-PXI) system, combined the Thermo Electric Cooler (TEC) with a duct heater, active heat dissipation intelligent control system is designed to fix the problem of heat dissipation in long distance in a narrowly closed environment without convection. According to the amount of thermal source and thermal energy, the smart control system can help to adjust the rate of heat dissipation by taking advantage of an adaptive control so that the performance of heat dissipation may be accumulated by its numbers. Last but not least, compared the traditional analog circuit controller with adaptive wavelet neural network controller, the research proves that its proposed active heat dissipation intelligent control system can reach an excellent and accurate temperature control. Speaking more precisely, adaptive wavelet neural network controller can be easily adaptive to any environment. It is equipped with a good capability of tracking and searching; and in terms of the effect of temperature control, it never actually jitters due to an input of voltage saturation compared with traditional analog circuit controller. All these can make chips able to adjust its adaptive rate of heat dissipation in accordance with the thermal source of the chips in a narrowly closed environment without convection.

Application of Particle Swarm Optimization Based on Neural Network for Artillery Range Prediction

Fri, 06 Mar 2026 04:06:00 GMT

title: Application of Particle Swarm Optimization Based on Neural Network for Artillery Range Prediction abstract: The firepower of artillery is one of main factors to influence the war effectiveness. Traditionally, the army utilizes the firing table to modify the artillery range, but the fabrication of firing table of artillery costs a lot of time and ammunition. In this study, some firing data of artillery are utilized to train the back-propagation neural network for artillery range prediction. Particle swarm optimization is utilized to increase the training speed of neural network and avoid getting stuck in local extreme. Besides, the orthogonal array is used to decrease the requirement of firing data and the proposed method is compared with the traditional back-propagation neural networks. Simulation results verify that the proposed method can not only increase the training speed of neural network but also have the satisfied performance of range prediction, and the mean absolute percentage error can approach to 1.173%. The proposed method in this paper is usable for artillery range prediction and feasible for application in the army.

IMM estimator based on fuzzy weighted input estimation for tracking a maneuvering target

Fri, 06 Mar 2026 04:05:54 GMT

title: IMM estimator based on fuzzy weighted input estimation for tracking a maneuvering target abstract: The application of target motion models and filters for interactive multiple model (IMM) estimator determines the effectiveness of maneuvering target tracking. In this paper, the fuzzy logic theory is utilized to construct the fuzzy weighting factor to improve the input estimation method and that is used to compute the unknown acceleration input for the modified Singer acceleration model. The proposed IMM estimator is operated mainly by two different target motion models combined with filters and the switch of target models is through the Markov transition probability matrix. The constant velocity model is combined with Kalman filter for the uniform target state estimation and the other one uses the modified Singer acceleration model to track the maneuvering target by the fuzzy weighted input estimation method. The performance of the proposed algorithm is verified by two different scenarios and compared with two IMM estimators. The target motion state of simulation condition contains the constant velocity, weak acceleration and strong acceleration. The simulation results show that the proposed IMM estimator has the better estimation precision in terms of tracking error. The modified Singer acceleration model combined with the fuzzy weighted input estimation method can track the maneuvering target effectively.

The Study to Apply Fuzzy Weighted Input Estimation for the Prediction of Target Trajectory in a Fire Control System.

Fri, 06 Mar 2026 04:05:51 GMT

title: The Study to Apply Fuzzy Weighted Input Estimation for the Prediction of Target Trajectory in a Fire Control System. abstract: The fuzzy weighted input estimation (FWIE) is proposed in this paper to solve the problem of noise disturbance and combined with the three-dimensional motion equation of target trajectory to construct the tracking rule of fire control system. FWIE can estimate effectively the input data of maneuvering target acceleration to obtain the precise target state and solve the problems from the traditional Kalman filter which can’t compute the precise estimation of target state because of the input information in the system. Simulation results show that FWIE can estimate the change of target state rapidly and precisely compared with the extended Kalman filter and the proposed tracking rule can improve the fire control system to figure out the target intercepting points with shorter miss distance.

Maneuvering target tracking using LOS and fuzzy input estimator

Fri, 06 Mar 2026 04:05:47 GMT

title: Maneuvering target tracking using LOS and fuzzy input estimator abstract: The performance of maneuvering target tracking relies on the extraction of useful information about the target state from the observations and the key factors focus on the suitable target models and the filters. A usable maneuvering target tracking technique is proposed in this paper. The line-of-sight (LOS) dynamic equations of interceptor–target are modified for the target model and the azimuth angle, elevation angle and the relative range between the radar and the target can be detected by the radar system fixed on the ground. A fuzzy input estimator based on Kalman filter and input estimation method is developed in which the fuzzy logic theory is used to accelerate the estimator tracking speed to the signal input. Three different filters are used to combine with the modified LOS target model for the verification and comparison. Besides, the proposed algorithm is compared with two different target tracking techniques. Simulation results prove the modified LOS target model is practical and the developed fuzzy input estimator can increase the estimation precision. The proposed algorithm has the least estimation error of target position compared with the other two different target tracking techniques.

Aiming Point Guidance Algorithm Based on Proportional Navigation Guidance Scheme

Fri, 06 Mar 2026 04:05:44 GMT

title: Aiming Point Guidance Algorithm Based on Proportional Navigation Guidance Scheme abstract: The proposed missile guidance algorithm is developed based on proportional navigation guidance scheme and the computation of aiming point. This research calculates the line-of-sight rate and the position of aiming point according to the current dynamics of missile and target, and applies particle swarm optimization to optimize and update the navigation constants of proportional navigation guidance continuously to figure out the missile control commands of lateral acceleration. Therefore, the missile will be guided to the aiming point as the computed target collision position. Simulation experiments prove the proposed guidance algorithm has the satisfied interception performance in a three dimensional engagement space with noise disturbance. The proposed method uses the shorter miss distance and less interception time compared with the proportional navigation guidance law and this could reduce the energy consumption as well. The outstanding guidance ability would be more obvious for intercepting the high agility aircraft. Furthermore, a novel artificial intelligence missile guidance algorithm is reproduced to execute the simulation experiments to compare the guidance technique with the proposed method. The proposed guidance algorithm is feasible to be applied to the real missile guidance system due to the advantages of simplicity and robustness.

Robust self-adaptive Kalman filter with application in target tracking

Fri, 06 Mar 2026 04:05:41 GMT

title: Robust self-adaptive Kalman filter with application in target tracking abstract: Kalman filter has been applied extensively to the target tracking. The estimation performance of Kalman filter is closely resulted by the quality of prior information about the process noise covariance (Q) and the measurement noise covariance (R). Therefore, the development of adaptive Kalman filter is mainly to reduce the estimation errors produced by the uncertainty of Q and R. In this paper, the proposed self-adaptive Kalman filter algorithm has solved the problems of covariance-matching method about the determination of the width of the window and the addition of storage burden and that can update Q and R simultaneously. Simulation results confirm that the proposed method outperforms the traditional Kalman filter and has the better estimation performance than the other two adaptive Kalman filters in the target tracking. The developed filtering algorithm has the following characteristics: high robustness, low computing load, easy operation and tuning Q, R simultaneously.

Dynamic interception point guidance algorithm based on particle swarm optimization

Fri, 06 Mar 2026 04:05:38 GMT

title: Dynamic interception point guidance algorithm based on particle swarm optimization abstract: The engagement of target-interceptor is an extremely complicated and nonlinear problem. Most literatures of developed guidance algorithms are hard to work in real-time missile guidance systems because of the complicated design of controllers, restriction in specific condition or excess computing loading. In this paper, the proposed guidance algorithm computes the predicted interception point of target-interceptor and applies particle swarm optimization to optimize the lateral acceleration control commands of missile where the definition of fitness function can guide the missile toward the predicted interception point when the computed fitness value is the minimum. According to the results of simulation experiments, the proposed method has the satisfied target-kill performance to the superior aircraft with high agility. The missile can greatly revise the flight route toward the computed collision course at the initial pursuit stage and the course curve of missile is flatter than the other two guidance laws. Besides, the proposed method can reduce the occurrence of big lateral acceleration control commands acting on the missile to avoid unlocking the evasive target at the terminal stage. As a result, the proposed guidance algorithm based on particle swarm optimization is very effective without using the complicated nonlinear control methods and excess storage burden of computer. It is a simple and feasible missile guidance algorithm due to the advantages of simplicity and effectiveness just like the proportional navigation guidance law but the performance of proposed guidance algorithm is better than proportional navigation guidance law and the other guidance algorithm designed by particle swarm optimization.

Apply Optimization Algorithm to Develop Parallel Navigation Guidance Law

Fri, 06 Mar 2026 04:05:33 GMT

title: Apply Optimization Algorithm to Develop Parallel Navigation Guidance Law abstract: The proposed missile parallel navigation guidance law utilizes the current dynamics of both the target and the missile to compute the line-of-sight rate and applies missile acceleration control command equations within the framework of proportional navigation guidance to control the missile's flight route. In this study, particle swarm optimization is used to continuously optimize the navigation constants and update the missile acceleration control commands, where the line-of-sight rate is defined as the fitness function of particle swarm optimization. The missile moves toward the target in parallel navigation guidance mode when the computed value of the fitness function approaches zero during the optimization process. There are three different pursuit-evasion scenarios in a three-dimensional space in the simulation experiments, where the target information includes noise disturbances to better simulate real target-interceptor engagement conditions. The simulation results prove that the proposed guidance theory is highly effective in intercepting maneuvering targets with high G-force, whereas proportional navigation guidance cannot achieve this effectively. Furthermore, a missile guidance algorithm using particle swarm optimization has been reproduced, which serves as the main motivation for developing the proposed guidance method in this study. The experiments demonstrate that the PSO-based missile guidance algorithm may produce oscillations in the control commands during pursuit-evasion, which could affect both the guidance performance and structural integrity of the missile. However, the proposed guidance method in this study not only mitigates these oscillations but also achieves better overall guidance performance.

Imaging of Rough Surfaces by Near-Field Measurement

Thu, 05 Mar 2026 04:08:25 GMT

title: Imaging of Rough Surfaces by Near-Field Measurement

Stochastic Nature of Voltage-Controlled Charge Dynamics in AlOx Magnetic Tunnel Junctions

Thu, 05 Mar 2026 04:08:20 GMT

title: Stochastic Nature of Voltage-Controlled Charge Dynamics in AlOx Magnetic Tunnel Junctions abstract: Spintronic memristors based on ferromagnetic metal/oxide heterostructures have recently enabled reversible manipulation of both magnetic properties and resistive switching (RS), offering promising prospects for multibit memory and neuromorphic computing. In this study, we investigate the stochastic nature and relaxation processes of charge dynamics induced by localized oxygen vacancy (VO) in AlOx-based magnetic tunnel junctions (MTJs). We observe that random telegraph noise (RTN) exhibits charge stochasticity at specific bias voltages in the low resistance state (LRS), reflecting the competition and transition between charge capture and emission states against the thermal energy. This behavior reveals that the thermally unstable charge stochasticity originates from localized traps in the AlOx barrier. In contrast, the high resistance state (HRS) favors the RTN emission states, indicating the dominance of direct tunneling effects. Through numerical calculations based on the tight-binding (TB) model and experimental results, we demonstrate that voltage-driven shifts in the VO position within the AlOx barrier, associated with RS, govern the charge dynamics of the MTJs investigated. These findings provide valuable insights and practical implications for the development of next-generation devices leveraging charge stochasticity in AlOx-based MTJs.

Hydrogen-Sensitive Antisymmetric Magnetoresistance in Co/Pd Multilayers Driven by Anomalous Hall Effect and Domain Wall Motion

Thu, 05 Mar 2026 04:08:12 GMT

title: Hydrogen-Sensitive Antisymmetric Magnetoresistance in Co/Pd Multilayers Driven by Anomalous Hall Effect and Domain Wall Motion abstract: We report tunable antisymmetric magnetoresistance (MR) in Co/Pd multilayers, governed by the interplay among current direction, magnetization orientation, domain wall dynamics, and hydrogen absorption. Under ambient conditions, the presence of perpendicular magnetic anisotropy (PMA) and domain wall motion gives rise to a pronounced antisymmetric MR. Through a combination of magneto-optical Kerr microscopy and magnetotransport measurements, we attribute this behavior to the anomalous Hall effect (AHE) inherent in the PMA state. Upon hydrogen exposure ranging from a vacuum of 1 × 10–3 mbar to 1 bar H2, the magnetization of the Co/Pd multilayers undergoes a spin reorientation transition, progressively tilting from the out-of-plane direction toward the in-plane orientation, as evidenced by a reduction in remanence from 100% to nearly 20%. This reorientation is accompanied by a pronounced shift of approximately 50 Oe in the MR spike, observable even under low H2 pressures up to 40 mbar. At higher hydrogen pressures approaching 1 bar, the AHE signal decreases by more than 70%, while the asymmetric MR spikes under a perpendicular magnetic field diminish from 0.1% to nearly 0.0%. In contrast, symmetric MR spikes of about 0.05% appear under in-plane magnetic fields, confirming the emergence of in-plane anisotropy. These findings demonstrate the pronounced hydrogen sensitivity of the MR spike shift, amplitude, and symmetry in Co/Pd multilayers, establishing a controllable platform for tuning spin-dependent transport with promising potential for multifunctional spintronic sensing applications.

High-Sensitivity Magnetoresistive Sensor With Optimized Orthogonal Exchange Bias for Low-Field Measurement

Thu, 05 Mar 2026 04:08:05 GMT

title: High-Sensitivity Magnetoresistive Sensor With Optimized Orthogonal Exchange Bias for Low-Field Measurement abstract: An MTJ-based magnetoresistive sensor that combines optimized orthogonal exchange bias with a previously proposed magnetic flux guiding structure is studied for improved overall performance. Due to the orthogonal exchange bias, the magnetoresistive sensor exhibits reduced hysteresis, which results in improved reversibility and linearity. On the other hand, high sensitivity is still achievable owing to the magnetic flux guiding structure and double-staged magnetic flux concentrators (MFCs). At room temperature, with double-staged MFCs, the sensitivity of the studied sensor is as high as $4.66\times 10^{4}$ %/mT at zero field, and the detectivity is about 23 pT/Hz $^{1/2}$ at 1 Hz and about 2.3 pT/Hz $^{1/2}$ at 100 Hz. The improved overall performance demonstrates great potential for application of the studied sensor in low-field measurement with high sensitivity at room temperature.

A low-temperature variation reference current source with digital counting auto-calibration scheme

Thu, 05 Mar 2026 04:07:58 GMT

title: A low-temperature variation reference current source with digital counting auto-calibration scheme

Development of an Intelligent Road Pothole Analysis System: Integrating YOLO Recognition and Large Language Models for Improvement Suggestions

Fri, 27 Feb 2026 04:07:09 GMT

title: Development of an Intelligent Road Pothole Analysis System: Integrating YOLO Recognition and Large Language Models for Improvement Suggestions

Development of Vacuum-Chamber-Type Capacitive Micro-Pressure Sensors

Fri, 27 Feb 2026 04:05:14 GMT

title: Development of Vacuum-Chamber-Type Capacitive Micro-Pressure Sensors abstract: This study presents the development of a capacitive pressure sensor tailored for measuring the dynamic pressure of flow fields. The sensor is fabricated using the UMC 0.18 μm CMOS-MEMS process, incorporated with additional post-processing steps such as metal wet etching, supercritical CO2 drying, and parylene encapsulation. The sensing architecture employs AD7746 as a capacitance-to-voltage converter (CVC), enabling the conversion of capacitance signals into voltage outputs for enhanced measurement fidelity. Structurally, the capacitive pressure sensor features a vacuum-sealed diaphragm capsule design with dual movable circular membranes functioning as sensing electrodes. A contact-mode capacitive configuration with a trapezoidal or Gong-like vacuum-chamber diaphragm is adopted to improve linearity and sensitivity. The output sensitivity was determined to be feasible for measuring dynamic pressure at 1–2 Pa resolution.

Transformer-based Model for Spatial-Temporal Feature Learning and Action Recognition of Human Skeleton Data, Jiann-Chyi

Thu, 26 Feb 2026 04:08:33 GMT

title: Transformer-based Model for Spatial-Temporal Feature Learning and Action Recognition of Human Skeleton Data, Jiann-Chyi

A dual-loop CDR circuit with dual-mode capacitor multiplier-based loop filter and active inductor ring VCO

Fri, 12 Dec 2025 04:05:23 GMT

title: A dual-loop CDR circuit with dual-mode capacitor multiplier-based loop filter and active inductor ring VCO abstract: This paper proposes A half-rate dual-loop clock and data recovery (CDR) circuit with dual-mode capacitor multiplier-based active loop filter and active inductor ring voltage controlled oscillators (VCO). The proposed capacitor multiplier-based active filter allows for changing the operational mode of the CDR circuit to adjust the loop bandwidth, thereby enhancing stability and reducing required area. The improved active inductor ring VCO features a wide tuning range and low power consumption. The proposed CDR is implemented in the 90 nm CMOS technology. The simulation results show that the rms jitter and the peak-to-peak jitter of the recovered data is 5.3ps and 31.1ps at 2.7-Gb/s data rate. For the recovered clock, the rms jitter and peak-to-peak jitter are 5.1ps and 28ps at 1.35 GHz clock frequency. This work consumes 11.2mW with 1.2V power supply.

C++從零開始：PBL實戰×問題導向程式設計

Wed, 19 Nov 2025 05:47:47 GMT

title: C++從零開始：PBL實戰×問題導向程式設計

Electromagnetic Inverse Scattering of Anisotropic Objects Using Deep Learning

Thu, 16 Oct 2025 06:22:39 GMT

title: Electromagnetic Inverse Scattering of Anisotropic Objects Using Deep Learning abstract: 電磁成像涉及多重散射與高度非線性的反演問題，而傳統研究多依賴正則化演算法求解。雖然此類演算法在一定程度上能提升成像穩健性，但其高昂的計算成本與長時間的運算需求，限制了即時成像的實現。此外，在抑制多重散射與處理高度非線性問題方面，仍存在技術瓶頸。本論文提出一種先進的人工智慧以改善電磁成像的反演問題，並透過整合深度學習和注意力機制，不僅有效提升重建影像的解析度，還能顯著降低計算時間，實現即時電磁成像的目標。我們透過發射TM(Transverse Magnetic)波與TE(Transverse Electric)波來照射各向異性物體，並接收其散射場資訊。接著，透過馬克斯威爾方程式(Maxwell’s Equations)與格林函數(Green’s Function)的數值計算方法來蒐集與分析量測的散射場資訊。在數值模擬中，由於各向異性物體的介電張量(Permittivity Tensors)較為複雜，進一步增加了非線性與不適定性的問題。為簡化訓練過程的複雜度，我們利用量測所得的散射場資訊經由反向傳播法(Back Propagation Schemes)進行初步介電係數的估測。然而，初步介電係數與真實介電係數仍存在一定誤差，因此本論文開發了三種不同的類神經網路架構，以提升影像重建的準確度與解析度。第一部分，我們開發了一種卷積類神經網路(Convolution Neural Network, CNN)結合混合注意力機制(Hybrid Attention, HA)來克服電磁成像的反演問題。混合注意力機制架構由通道注意力機制(Channel Attention, CA)與空間注意力機制(Spatial Attention, SA)組成。CA透過動態調整特徵圖不同通道的重要性，以強化關鍵特徵的影響權重；而SA則學習影像中不同區域的重要性，透過自適應方式分配像素之間的權重，提高重建影像的精確度與解析度。第二部分，我們首先開發了一種感知生成對抗網路(Perceptual Generative Adversarial Network, PGAN)，並結合自注意力機制(Self-Attention)來增強影像細節與準確性。相較於傳統的生成對抗網路(GAN)，PGAN在損失函數中額外引入感知約束(Perceptual Constraint)，進一步優化影像品質。此外，自注意力機制能夠捕捉遠距離特徵之間的關聯性，進一步提升重建影像的效果。最後，我們提出了一種點對點串接式類神經網路架構用以細化重建的電磁影像。與前述兩種網路架構不同，該方法每個級聯階段可以與其他級聯階段共享權重，每個級聯都會優化前一個級聯的輸出以間接地獲得高對比度的介電係數分布。相較於傳統的類神經網路，點對點串接式類神經網路從輸入到輸出都是聯合訓練的且可以同時優化所有階段。這種統一的訓練策略使模型能夠以全域一致的方式進行學習。在數值模擬中，我們分別加入5%和20%的雜訊以驗證所提方法的有效性。結果顯示，第一部分的CNN在加入混合注意力機制後，亦明顯改善影像品質；第二部分的PGAN在結合自注意力機制後，顯著提升了重建影像的解析度；第三部分提出之點對點串接式類神經網路架構，相較於傳統的U-Net結構，具備更優異的影像重建能力。透過逐層優化前一級輸出結果的方式，能有效提升影像的解析度與對比度，進而強化重建品質與細節表現。

MOSA: Matrix Optimized Self-Attention Hardware Accelerator for Mobile Device

Fri, 10 Oct 2025 04:05:29 GMT

title: MOSA: Matrix Optimized Self-Attention Hardware Accelerator for Mobile Device abstract: The Self-Attention mechanism, which lies at the core of Transformer architectures, plays a vital role in capturing long-range dependencies. However, its high computational complexity and significant memory requirements pose major challenges for resource-constrained hardware such as mobile devices. In particular, frequent memory accesses and inefficient matrix multiplication operations often result in performance bottlenecks. Therefore, the development of dedicated hardware accelerators for Self-Attention, focusing on optimizing matrix operations and reducing memory usage, is essential for improving AI processing efficiency on mobile devices. This paper presents a Self-Attention hardware accelerator designed specifically for mobile devices. By optimizing the matrix multiplication process, the accelerator effectively reduces data transmission and memory access frequency, thereby lowering the overall computational complexity. It also reuses intermediate computation results, minimizing frequent memory read and write operations, which significantly reduces memory bandwidth requirements. This data reuse strategy not only cuts down on redundant computations but also greatly enhances computational efficiency. In addition, the accelerator eliminates the need for Key Transpose operations, further simplifying certain computational steps. Compared to traditional algorithms, the idle time is reduced by 66.5%., and signal line transmission costs are reduced by approximately 80%. With an optimized hardware architecture, mobile devices can efficiently support complex deep learning applications, such as speech processing and image recognition.