In order to solve the problem that traditional telescopic systems cannot simultaneously achieve performance requirements such as large field of view and lightweight, the design method of an off-axis three-mirror afocal optical system based on free-form surfaces was researched. Starting from the basic quadratic surface on-axis three-mirror afocal structure, the initial structure of the off-axis three-mirror afocal system was obtained through aperture off-axis processing, and system optimization was carried out on this basis. Introducing Zernike polynomial free-form surfaces to replace traditional quadratic surfaces can further improve imaging quality while reducing system volume and achieving miniaturization requirements. Finally, an off-axis three-mirror telescopic structure with an aperture of 200 mm, a magnification of 5, and a field of view angle of 3° was obtained by optimization. The MTF value in the full field of view of the system is close to the diffraction limit. An analysis of the image quality of the system after the introduction of free-form surfaces was conducted, and the results showed that when only two mirrors were introduced into the free-form surface, the system had already reached the diffraction limit, which can reduce the processing difficulty to a certain extent.

A zoom optical system for thin film damage threshold test is designed to change the size of the optimal focusing spot and the distribution of energy density by zooming, which solves the phenomena of gas explosion and “internal sculpture” in the thin film damage threshold（LIDT）test caused by the focusing lens in the defocusing mode due to the high energy of the laser. Experiments show that the focusing optical system designed in this paper, the optimal focusing spot size can be adjusted by zoom. When the pitch changes from 18.152 mm to 6.851 mm, the effective focusing spot diameter changes from 100 m to 700 m. The beam morphology meets the Gaussian distribution. The laser energy distribution in line meets the use of the requirements. The reasonableness and validity of the focusing optical system are verified. The focusing optics can be adjusted by zoom. The system uses zoom to change the focal spot size and energy distribution. The best focusing spot acts on the surface of the film system. The formation of laser damage inside the component is prevented. The probability of misjudgment of the threshold damage of the film is reduced. The accuracy of the laser damage threshold test is improved.

The effects of core error in scanning and lensless flexible optical endoscopic phase imaging based on coherent fiber bundle（CFB）were studied. Utilizing wave optics theory, a theoretical model has been developed for scanless and lensless flexible optical endoscopic 2D phase imaging based on CFB, addressing transmission mode and core length errors in CFB. The results show that the imaging resolution is different between single-mode transmission and multi-mode transmission, and the phenomenon of super diffraction limit resolution exists in multi-mode transmission. Core length error has a serious impact on imaging. When the core length error is large, imaging cannot be done. When the core length error is small, the imaging signal-to-noise ratio decreases. Wiener filter can be used for fuzzy images with low signal-to-noise ratio affected by small core length error. After several iterations, the signal-to-noise ratio of images can be increased by more than 5 dB, and the normalized root mean square error of images is less than 0.5, which effectively improves the image quality. Therefore, in order to achieve imaging in the scanless and lensless flexible optical endoscopic phase imaging system based on CFB, it is necessary to either use single-mode CFB or to suppress multi-modes in multi-mode CFB, thereby ensuring single-mode operation.

Quantum dot photodetector is a hot spot in current research, which has a wider spectral response range（350~2 000 nm）due to the use of special photosensitive materials. However, in the process of imaging chip preparation, because of the influence of the immature synthesis process and the low image element transplantation rate, the obtained images will have the image problems of blind flash element bunching and blurred details. Therefore, in this paper, in order to improve the quality of the image and attenuate the effects of the blind flash element noise and detail blurring in the image, we propose to enhance the wide-spectrum image with an adaptive median filtering algorithm and an inverse sharpening mask algorithm, and take advantage of the pipeline processing of the FPGA（Field Programmable Gate Array, FPGA）, which is deployed on the Xilinx A7 series hardware deployment. The results show that the peak signal-to-noise ratio of the processed image can be increased by 17% when a frame with a resolution of 640×512 is input. With a clock of 50 MHz, the processing takes about 6.16ms, which is of practical significance for the optimization of wide-spectrum imaging systems with high real-time performance and low hardware occupancy.

In order to solve the problems of limited parameter redundancy deployment and limited detection speed in the study of infrared anti-interference in deep learning, this paper proposes a lightweight method based on improved YOLOv5s, which uses Dy-MobileNetV3 to replace the backbone part of YOLOv5s. Dy-MobileNetV3 replaces the SE channel attention mechanism in the lightweight network MobileNetV3 with the CA coordinate attention mechanism to improve the accuracy and computing efficiency, and uses the dynamic convolution module to reconstruct the MobileNetV3 network to maintain the lightweight computing cost and improve the accuracy. Adaptive label smoothing is introduced to reduce the overfitting of the model and improve the generalization ability of the model. Compared with the basic model YOLOv5s, the proposed DyM-YOLOv5s model reduces the number of parameters by 85%, the image processing speed is 3.5 times that of the original, and the accuracy is increased by 4%.

Conventional intensity modulation and direct detection passive optical networks（IMDD PONs）have limited flexibility to cost-effectively accommodate a large number of emerging applications with largely diversified requirements in terms of bandwidth, latency and connectivity. To effectively address this technical challenge, a novel Inverse Fast Fourier Transform/Fast Fourier Transform（IFFT/FFT）-enabled Point-to-Multipoint（P2MP）optical transceiver has been recently proposed and experimentally validated in upstream IMDD PONs. However, comprehensive investigations of the flexibility of this technique are still required to validate its suitability for future applications. To this end, by using presentative upstream IMDD PON simulation systems, this paper first identifies the optimum parameters and then investigates the transceiver’s flexibility and resulting upstream transmission performances. Simulation results indicate that in cases where the optical fiber channel attenuation is not significant, the fiber transmission distance can be flexibly adjusted without significantly reducing the system transmission capacity. In addition, each ONU can flexibly change its channel count without significantly reducing transmission capacity if the channel counts are≤10. Furthermore, by dynamically changing each ONU bitrate, each ONU can flexibly alter its ONU launch power dynamic range.

The control of DC voltage synchronization has always been a major challenge in microgrid systems, as it directly affects the communication and energy transfer efficiency between microgrids. This article proposes a photovoltaic new energy based photovoltaic voltage control method for DC microgrids, aiming to improve the synchronization performance of DC voltage. By adopting a distributed collaborative control strategy and optical communication isolation technology, a linear optical communication secondary regulator based on integration is introduced to achieve synchronous control of output voltage and derivative. By introducing global performance indicators, voltage ripple and steady-state error were effectively reduced, and their effectiveness was verified in experiments. The voltage synchronization error of the local microgrid was reduced. The synchronization error of adjacent microgrids was minimized. The interaction between DC-DC buck converters were transfered from the physical layer to the communication layer. The simulation experiment results show that under input voltage disturbances, this method has good voltage synchronization performance, voltage ripple elimination ability, and voltage convergence speed, highlighting the important role of optical communication technology in photovoltaic new energy DC microgrids.

Ionospheric delay is one of the important sources of error in precision single-point positioning, time synchronization and other related fields. Accurate prediction of the total ionospheric electron content is an important prerequisite for compensating for ionospheric delay. This paper uses the long short-term memory（LSTM）network prediction algorithm to construct the TEC prediction model. The influences of different hidden layers, the number of neurons, the number of training data, the number of training times, and whether the data is preprocessed on the prediction of ionospheric data are verified. The best parameters of ionospheric prediction are obtained. The results show that the number of hidden neurons in the 2nd layer, the 1st and 2nd layers are 200 and 300, respectively, the number of input neurons is 168, the number of output neurons is 12, and the number of model iterations is 400 times, the prediction effect is the best, and the root mean square error of the prediction results of 18 single-site TEC prediction results is 43.87, which is 275.58 lower than that of the BP neural network algorithm. The LSTM network prediction algorithm effectively improves the prediction accuracy, and the clock error can be effectively compensated.

Circular fringe projection profilometry（CFPP）, as a single-shot measurement method, can calculate the absolute phase from single wrapped phase distribution. However, in the existing CFPP based on Fourier transform, the fringe pattern needs to be transformed into the polar coordinate system to acquire the phase, and then the phase value needs to be transformed back into the Cartesian coordinate system. The coordinate transformation and the used interpolation operation increase the computing load and introduce some unexpected errors. To directly extract the phase information from the circular fringe, this paper investigates circular fringe projection profilometry based on windowed Fourier transform. One of the advantages is that it can avoid the influence on the reconstruction from phase errors around the center and reduce the difficulty of phase unwrapping by locating the circular center at edge region of the field of view. Furthermore, the proposed method can obtain the absolute phase from single wrapped phase by calculating the fringe order difference between the position of the circle center and the selected unwrapping starting point. In addition, in order to improve the accuracy of calculating pixel displacement amount, which relates to the height of the measured object, the inverse fringe projection technique is introduced into our method to correct the period of the reference fringes to be a constant. Both simulations and experimental results demonstrate that the circular fringe projection based on windowed Fourier transform can achieve measurements of off-plane objects, with a root mean square error less than 0.05 mm for planar and object measurements. The research work fundamentally ensures the measurement accuracy of circular fringe projection profilometry, offering advantages and reference value for fast and high-precision 3D measurements of objects in out-of-plane motion.

In order to effectively identify the characteristic parameters of hemispherical shell resonator, an extended Kalman filter algorithm is proposed to identify the four characteristic parameters of hemispherical shell resonator（HSR）, which includes anisotropy of stiffness , primary axis of stiffness , anisotropy of damping (1/ ) and primary axis of damping . In this paper, based on the amplitude control equation and orthogonal control equation of the non-ideal hemispherical shell resonator, the parameter identification equation based on the extended Kalman filter algorithm is constructed, and the simulation model of the hemispherical resonator gyro（HRG）is built, so as to verify the effectiveness of the proposed method. The simulation results show that all the characteristic parameters converge to the set expected value within 15 seconds. The proposed method can effectively identify the characteristic parameters of the hemispherical shell resonator, and can provide a favorable guide for the error calibration, compensation and system accuracy improvement of the hemispherical resonator gyro.

Benefiting from the time-frequency analysis and multi-resolution characteristics of wavelet transform, wavelet transform profilometry（WTP）can accurately extract the phase map carrying the height information of the measured surface from a single frame of fringe pattern. The difficulty and accuracy of phase demodulation in wavelet transform profilometry depend on the choice of wavelet. Complex wavelets can be directly used for fringe phase demodulation. The nth-order derivatives of the one-dimensional Gaussian function satisfy the admissibility condition of wavelets, and the spectral distribution of each side-band exhibits smooth and asymmetric characteristics. These 1D functions, as 1D real wavelet bases, combined with the 2D complex wavelet framework, can be used to construct two-dimensional complex wavelets with good directionality and smoothness, suitable for fringe analysis, effectively improving the accuracy of phase calculation in wavelet transform profilometry. The spatial-frequency characteristics of 2D complex wavelets, constructed from the first, second, and third derivatives of a one-dimensional Gaussian function, are thoroughly examined through theoretical analysis and numerical calculations. These wavelets are then applied to the analysis of low signal-to-noise ratio and low-contrast fringes to reconstruct the 3D surface profile of the object. Using Phase Measuring Profilometry（PMP）results as a reference, the standard deviation（STD）of 3D surface reconstruction errors, based on the three 2D complex Gaussian wavelets developed in this study, remains consistently below 0.07 in planar measurements. For non-uniform reflective surfaces with low signal-to-noise ratio fringes, the 2D WTP demodulation method, employing the proposed wavelets, outperforms both the complex Mexican hat wavelet and the traditional Fan wavelet in terms of reconstruction accuracy. In non-dark-field regions, the reconstruction error STD is maintained below 0.05, effectively enhancing measurement precision. This research not only enriches the wavelet basis function library for 2D WTP but also expands its potential applications in optical measurement and fringe pattern analysis.

Transparent object detection is a difficult problem in the field of machine vision. Transparent objects can be rapidly detected by using multi-view light field information and polar space constraints. However, such algorithms often have low efficiency due to the large number of feature points and high similarity. To solve the above problems, the adaptive density clustering method is used to filter the feature points set, and the global and local feature significance and motion consistency constraints are combined to ensure the accuracy of feature matching and improve the speed of the algorithm. Furthermore, the slope difference between the horizontal and vertical dimensions of the feature information in the polar space domain is used to realize the target detection, which reduces the fitting process and realizes the rapid detection of transparent objects. The results show that compared with other similar algorithms, the algorithm proposed in this paper can improve the operation speed by more than 5 times while ensuring the detection accuracy.

In order to improve the sensitivity and accuracy of the eddy current sensor, the influence of the excitation frequency and the structural parameters of the probe coil were studied on the performance index of the eddy current sensor. Based on the basic theory of Maxwell electromagnetic field and the finite element analysis method, the equivalent model of the eddy current sensor was simulated on the COMSOL simulation software. The simulation results show that the excitation frequency has the greatest effect on the eddy current distribution. The increase of the excitation frequency increases the eddy current intensity on the surface of the measured conductor, but the detection depth decreases rapidly. When the inner diameter and outer diameter of eddy current probe coil are larger, the thickness is smaller, the sensitivity of eddy current sensor is higher. The research results provide important reference value for the selection of excitation frequency of eddy current sensor in voice coil fast reflection mirror and the optimization design of probe coil structure parameters.

With the rapid development of modern medical technology, electronic endoscopes are applied to the diagnosis and treatment of diseases in the body, and the accurate measurement of their optical imaging performance is of great significance. This article introduces an improved MTF measurement device and its method based on manual detection of electronic endoscopes. The method uses direct detection of resolution plate imaging contrast comparison values to obtain the modulation transfer function that reflects the imaging ability of the optical system. The entire process is completed using automatic integrated image processing technology, which has become the core and effective main component of the optoelectromechanical integration system. The experimental data and the implementation effect installed on the production line reflect that the qualification rates of the manual inspection visual method and the detection method of this measuring device are 94% and 96%, respectively. The MTF curve of the 7.87~12.5 lp·mm-1 spatial frequency band is more consistent with the theoretical curve of the designed lens, which verifies the effectiveness of the method proposed in this paper. The use of the production helps to reduce human error and significantly improve production efficiency.