15 September 2025, Volume 45 Issue 5
-
Select all|
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 1-6. https://doi.org/10.13954/j.cnki.hdu.2025.05.001In order to construct generalized barycentric coordinates with non-negativity and locality, this paper proposes a method for constructing visible natural neighbor coordinates based on minimizing Dirichlet energy. By computing the second-order Voronoi diagram and visible vertices, the visible natural neighbors are obtained, ensuring the locality of the coordinates. Subsequently, the coucept of Dirichlet evergy is induced, and the smoothness of the coordinates is ensured by minimizing the Dirichlet energy. Numerical examples demonstrate that the proposed method can construct barycentric coordinates with non-negativity and stronger locality within any simple polygon.
-
Study on vector wave functions for electromagnetic scattering by two-dimensional bianisotropic mediaJournal of Hangzhou Dianzi University. 2025, 45(5): 7-13. https://doi.org/10.13954/j.cnki.hdu.2025.05.002This paper proposes a method for studying the cylindrical vector wave functions(CVWFs) of electromagnetic scattering by the two-dimensional bianisotropic media. Based on the two-dimensional CVWFs, the electromagnetic field in the two-dimensional bianisotropic media can be expanded into the integral expression of the CVWFs, while the scattering field and incident field outside the bianisotropic media can also be expanded into the form of the CVWFs in media. By using the continuity of the tangential component of the electromagnetic field on the surface of the bianisotropic media, the extended boundary condition method and the orthogonality of the CVWFs, the scattering coefficients of the scattering field in terms of CVWFs in free space can be derived, and then the radar cross section of the two-dimensional bianisotropic media can be obtained. Numerical results comparing with existing literature coincide enough, and discussions are drawn at the end.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 14-22. https://doi.org/10.13954/j.cnki.hdu.2025.05.003Based on the 40 nm CMOS process, a fractional-N phase-locked loop with frequency range of 19~20.25 GHz for frequency-modulation continuous-wave chirp is proposed. An improved pre-charged frequency-phase detector is used to alleviate the blind zone problem in the frequency-phase detector and quickly captures the frequency difference of the input signals for achieving fast locking in the phase-locked loop and meeting the requirements of slope rate of chirp. An improved multistage noise-shaping structure is used in the Σ-Δ modulator to effectively increase the length of the output sequence and reduce fractional spurs. Simulation results show that with an output frequency of 19 GHz, the phase noise of the phase-locked loop is -100 dBc@1 MHz, and the maximum modulation rate of the linear wave is 310 MHz/μs, which has much faster modulation speed by comparing with traditional phase-locked loop. The overall power consumption is 65-105 mW under 1.8 V supply voltage for VCO and 0.9 V supply voltage for the rest circuits.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 23-31. https://doi.org/10.13954/j.cnki.hdu.2025.05.004Predicting effluent quality is crucial for early warning and monitoring of sewage plant water quality, enabling dynamic regulation of sewage plants. However, industrial wastewater quality data often exhibit high volatility and nonlinearity. To address these challenges, this paper proposes a time-convolutional network(TCN) model based on the efficient channel attention(ECA) mechanism, bidirectional LSTM(BiLSTM), and CEEMDAN optimization. Firstly, the CEEMDAN decomposition technique effectively reduces the fitting difficulty of the input sequence. The ECA module captures local channel feature information, which the BiLSTM module extracts bi-directional temporal features and enables accurate prediction of industrial wastewater effluent quality time series by combining with the TCN network Finally, the output of each component is reconstructed. Experimental results demonstrate that the proposed TCN model achieves better accuracy in predicting the chemical oxygen demand COD of industrial wastewater effluent quality, with an R2 index of 0.90, surpassing other benchmark models. This study provides a basis for efficient dynamic regulation in industrial wastewater treatment plants.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 32-38. https://doi.org/10.13954/j.cnki.hdu.2025.05.005In order to meet the requirements of high precision and high reliability clock for time-consuming equipment, according to the characteristics of high long-term stability of global navigation satellite system(GNSS) timing and high short-term stability of crystal oscillator clock(COC), a scheme of taming crystal oscillator by GNSS second pulse is proposed, which realizes the complementary advantages of GNSS and COC, and improves the stability and accuracy of crystal oscillator output frequency. Based on the system of GNSS taming the constant temperature crystal oscillator, the singular value truncation algorithm is used to filter out the outliers in the system in real time; combined with the fading filtering algorithm, the system error is effectively reduced and the stable output is within a certain range. Experimental results show that the frequency accuracy of the constant temperature crystal oscillator can reach 3.4×10-12, and the timing accuracy is better than 20 ns.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 39-46. https://doi.org/10.13954/j.cnki.hdu.2025.05.006The impacts of anisotropic ocean turbulence on the propagation of anomalous vortex beams through an ocean channel are examined. Using Rytov theory, mathematical expressions for the orbital angular momentum mode detection probability and channel capacity of an anomalous vortex beam are derived. Numerical simulations are used to analyze the effects of light source parameters and anisotropic ocean turbulence parameters on the mode detection probability and channel capacity, which indicate that higher radial index values of the anomalous vortex beam result in better resistance to turbulence. In addition, an increase in most turbulence parameters, with the exception of the anisotropy factor and turbulent kinetic energy dissipation rate, leads to a decrease in the detection probability and channel capacity. Furthermore, the propagating performance of an anomalous vortex beam is compared with that of the Hankel-Bessel beam and the Laguerre-Gauss vortex beam in anisotropic ocean turbulence. It has been found that anomalous vortex beams have superior transmission abilities in anisotropic ocean turbulence under certain conditions.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 47-53. https://doi.org/10.13954/j.cnki.hdu.2025.05.007A high-gain low-noise amplifier operating in the X-band has been designed based on the 0.18 μm CMOS technology. The circuit employs a two-stage differential common-source structure and utilizes a cross-coupled capacitance neutralization to counteract the coupling effects from the Miller effect of the transistor gate-drain parasitic capacitance, and improves the circuit gain, stability, and reverse isolation. Transformer baluns are used for input, inter-stage, and output matching networks to achieve a compact impedance matching structure and conversion from single-ended to differential signal at ports. Testing results show that the low-noise amplifier has a 3 dB gain bandwidth of 9-11 GHz, a maximum gain of 25.1 dB, a noise figure of (3.6±0.1) dB, and a core chip area of 0.34 mm2 with a DC power consumption of 35 mW.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 54-61. https://doi.org/10.13954/j.cnki.hdu.2025.05.008In order to cope with the risk of equipment aging due to poor maintenance and the decline in profitability of equipment lessors, as well as the demand of equipment lessors to provide more additional services, by analyzing the demand for production equipment of mass customization products and the interaction between equipment service life and maintenance, an optimization model aiming to reduce equipment leasing costs and improve equipment utilization is established, in order to determine leasing, production, and maintenance plans that satisfy both the lessor and the lessee. A genetic algorithm-based discrete invasive weed optimization(GADIWO) is proposed for solving this model, and a mapping method for the variables is designed to make the algorithm applicable to discrete problems. It improves the diffusion mechanism by designing new crossover and mutation operators to enrich the diversity of solutions and avoid the algorithm from trapping into local optima. A large number of numerical experiments show that GADIWO's optimization ability is superior to classical invasive weed optimization and genetic algorithm under the same running time.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 62-69. https://doi.org/10.13954/j.cnki.hdu.2025.05.009In order to meet the requirements of lightweight and miniaturization of the U-shaped pan tilt in high-precision scanning leak locators, and improve the stability of the system, this paper proposes a multi-objective parameter optimization for the static structure of the U-shaped pan tilt based on response surface function and genetic algorithm. Using stress, deformation and moment of inertia as constraint functions, mass and volume as objective functions, multiple sets of models are designed and simulated to extract five response values and establish a response surface model. Subsequently, a genetic algorithm is used to analyze and optimize the results for optimizing the U-shaped pan tilt structure. Experimental results show that compared with the original prototype, the optimized model achieves a mass reduction of 4.30% and a volume reduction of 37.98%.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 70-76. https://doi.org/10.13954/j.cnki.hdu.2025.05.010The light field camera is a novel imaging device that has garnered considerable attention due to its ability to capture both light intensity and direction information in a single shot. Such capability holds immense promise for image processing applications. However, the current state-of-the-art super-resolution methods tend to overlook the critical occlusion relationship, leading to suboptimal reconstruction performance in complex occlusion scenarios. Moreover, prevailing approaches lack refinement in their exploration of structural and parallax information. To address these limitations, this paper proposes a novel multilevel feature fusion-based light field image super-resolution network. The network begins by extracting viewpoint maps from multiple viewing angles, allowing for a more comprehensive understanding of the occlusion relationship. Subsequently, by treating two diagonal views as a sequence, the network extracts gradient maps to mine crucial structural information. Finally, we convert the subaperture image into a lens images to extract valuable parallax information. Experimental results on open datasets demonstrate superior reconstruction performance of the proposed method, surpassing state-of-the-art approaches.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 77-86. https://doi.org/10.13954/j.cnki.hdu.2025.05.011Taking the active material of waste lithium battery cathode as the research object, a mild citric acid leaching system is selected to investigate the effects of acid concentration, reducing agent, solid-liquid ratio, reaction temperature, reaction time and other factors on cobalt leaching rate. Furthermore, the thermodynamics and the kinetics of cobalt leaching are analyzed. In terms of the selection of reducing agents, both the subsequent recovery of cobalt in the leaching solution (synthesis of cobalt ferrite) and the use of iron based reducing agents have achieved efficient cobalt leaching, while providing an ion foundation for subsequent cobalt recovery. Experimental results show that under the conditions of reaction temperature of 80 ℃, initial acid concentration of 1.25 mol/L, addition of reduced iron powder of 6.67 mmol/g, and solid-liquid ratio of 40 g/L, cobalt leaching rate can reach 99.95% after 120 minutes of leaching. In the range of pH<6.80 and (0.65~1.61)V＞E＞-0.31 V, reduction helps to accelerate the leaching of cobalt from the positive electrode active material of lithium batteries. The cobalt leaching process can be described by an unreacted core contraction model without a solid product layer. The reaction rate is controlled by a combination of surface chemical reactions and diffusion. The calculated apparent activation energy is 24.55 kJ/mol, and the expression for the prediction model of cobalt leaching rate in this system is obtained.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 87-92. https://doi.org/10.13954/j.cnki.hdu.2025.05.012A numerical method for solving fractional-order integro-differential equations with weak singularity is proposed based on neural network and Jacobi polynomials. Firstly, the existence and uniqueness of the solution of the equation are proved theoretically. Then, the equation is discretized based on the definition of Caputo fractional-order derivative and the Gauss-Jacobi numerical integration formula. Meanwhile, the neural network based on Jacobi polynomials is constructed to approximate the numerical solution of the equation. Finally, numerical experiments show that the method is effective.
-
Journal of Hangzhou Dianzi University. 2025, 45(5): 93-102. https://doi.org/10.13954/j.cnki.hdu.2025.05.013Explicit model predictive control (EMPC) moves the online computational burden of model predictive control (MPC) to offline computation by using multi-parametric programming which produces piecewise affine control laws defined over a set of polyhedral regions in the state space. This paper uses multi-level grids structure to store the polyhedral regions and their control laws which divides the state space into several levels and build the grids structure for each level. For the given state point, the method can quickly locate the grid and the responding polyhedral region and calculate the control input. Simulation results show that the proposed method can build the multi-level grid structure in rather short time and improve online searching effectiveness, especially for the control problems with high dimensional state space or a large number of polyhedral regions.
