In order to solve the restriction caused by inflexible parameter control and low analysis efficiency in the motor’s test system, this paper builds a multi-parameter measurement and control system. This paper introduces a system that can measure and test the main mechanical-electrical parameters of ultrasonic motors. Firstly, an overall scheme is designed according to the parameters and the input-output characteristics. Then in order to obtain the characteristics of variable preloads and temperature rapidly, an electric cylinder is employed as the source of pre-pressure and a temperature acquisition unit based on thin-film thermal resistance is designed; In view of the difficulties in flexible generation and real-time acquisition of high-frequency signals, a driving circuit which can satisfy the adjustment and measurement in multi-parameters is designed. An overall flowchart is formed for solving low efficiency and incompleteness occurring in the traditional test process. The flowchart especially includes synchronous acquisition processes that can measure transient characteristics of the stepwise movement. Experimental results show that the system covers the main 8 control parameters and 13 state parameters of the ultrasonic motor, and the test apparatus can achieve online accurate adjustments of control parameters and rapid measurement of state parameters. The efficiency in this system is 30% higher than the one in other traditional test system.This paper focuses on solving the problems of multi-parameter measurement and control and provides several effective strategies for dynamic modeling, characteristic evaluation and servo control of the ultrasonic motor.

In order to improve the environmental adaptability and energy conversion efficiency of ultrasonic motors, a novel piezoelectric and friction material with good temperature stability and low dielectric loss was been prepared and studied systematically in this paper. A new piezoelectric material was proposed based on the results of phase structure modulation, introducing of third end-member, introducing dipole defects to pin domain walls. In order to improve the energy transfer efficiency and environmental adaptability of the friction interface of ultrasonic motor, a novel polyimide composite material with high friction coefficient, low wear rate and high temperature resistance was designed. Finally, in the assembly of the ultrasonic motor, the piezoelectric and friction functional materials proposed in this paper are applied. The results show that the new piezoelectric ceramic materials have the advantages of low dielectric loss, high mechanical quality factor and high temperature stability compared with traditional materials. To be specific, the energy conversion efficiency of the ultrasonic motor increased by 3.3%. The new polyimide friction materials have significantly improved the friction coefficient, wear resistance and temperature range, which increased the maximum efficiency value of the ultrasonic motor by 6.19%. Moreover, after integrating the two new materials, the maximum efficiency value of the ultrasonic motor has increased by 13.6%, and the usage temperature range has increased from -40-70 ℃ to -60-120 ℃.The piezoelectric and friction functional materials proposed enhance the environmental adaptability of the ultrasonic motor and improve the output performance of the motor in this paper, which is of great significance for the application of ultrasonic motors in the equipment for aerospace and other high-end fields.

Preload force is a key factor for ensuring the frictional drive of the stator and rotor of an ultrasonic motor. In this paper, the influence of the preload force on the proportion of the contact antenna and driving area is first analyzed through simulation. The velocity fluctuation law of different preload forces is tested experimentally. The velocity and efficiency surfaces under different preload and torque are measured and plotted. With the help of the encoder and film temperature sensor, the changing process of the speed and stator interface temperature of the motor under different preload force was obtained. Based on the above analytical results, a preload force optimization criterion for different applications is proposed, and the ideal working range of the preload force is determined. With the growth of preload force, the contact area is synchronized with the drive area, while the speed of the driving area decreases and the speed fluctuation of the motor is somewhat reduced, the maximum efficiency of the motor occurs at a larger torque. Moreover, the highest rate of mechanical efficiency is at a certain preload force, temperature rise of the interface increases, and surface temperature rise fluctuates. According to the analysis results, the proposed optimization criterion determined that the ideal working preload force range for the TRUM60A ultrasonic motor is 260-320 N. The preload force in this range enables the ultrasonic motor to meet the requirements of low speed stability and small temperature rise, and achieve the ideal range of braking torque and mechanical efficiency.

In order to improve speed/position control accuracy and overcome nonlinearity accompanied by a multivariable coupling effect, the targeted speed and position control of the ultrasonic motor was proposed in this paper. Firstly, the control mechanism of the ultrasonic motor was demonstrated from the perspectives of continuous macroscopic motion and microscopic stepwise motion. Then, the sources of velocity fluctuation errors were clarified, and a multiparameter speed control model was established; the amplitude and frequency were integrated into the model by combining steady and dynamic links. In order to predict stepwise displacements, a two-stage second-order speed model was also constructed from the microscopic dimension. Then, the double-loop compound control algorithm was used to realize high-stability speed control, while high-resolution position control could be achieved through driving parameter optimization. Finally, a piecewise approximation strategy was adopted to combine the continuous and stepwise motion for high-precision positioning. The experimental results show that speed stability under dual-loop compound control is 0.44%, which is two times better than that of single-loop control. In terms of position control, the open-loop position resolution of the ultrasonic motor reached 0.375 μrad, and a positioning accuracy of 1.7 μrad is achieved. The proposed control strategy integrates the control characteristics of different parameters and loops, which effectively improves the speed and position control accuracy of the ultrasonic motor. This work lays the foundation for extended applications in advanced precision equipment.

To satisfy the requirements of micron-level measurement accuracy and second-level measurement time in industrial detection, this study proposes a full-field heterodyne short coherent topography measurement scheme. A light source with short coherence is developed to achieve long stride measurement and save scanning time. This study uses full-field heterodyne technology to achieve fast inversion of interference contours, and suppresses the effects of vibration and DC noise on measurement accuracy to improve the efficiency of data inversion. Furthermore, an experimental verification system was set up, and the detection time of the system was verified to be less than 10 s with measurement accuracy superior to 2 μm. The measurement time achieved by the system could be improved to less than 5 s with micron-level measurement accuracy by further optimizing the design. Therefore, the proposed technology was verified to be capable of satisfying the measurement requirements in industrial detection with the advantages of faster measurement speed and higher measurement accuracy compared to existing methods, and has prospective applications in the field of highly efficient industrial detection.

In dynamic light scattering technology, the influence of autocorrelation data noise on measurement results primarily depends on the particle-size inversion algorithm. In multiangle measurement, angular weighting becomes another important factor restricting the influence of noise on the measurement results.Based on the analysis of angle weighting mechanism of multiangle dynamic light scattering, the inhibition effect of the angle weighting method of light intensity average and the iterative recursive method on the measurement of noise is studied. The results demonstrate that, in the absence of noise, for the small particles of the unimodal distribution, the PSD is slightly broadened when weighted by the iterative recursive method, and for the medium and large particles, the peak value error increases slightly when the light intensity average method is adopted. With an increase in noise, the performance index of iterative recursive inversion exhibits no obvious change, while the peak value error and PSD error of the results obtained via the light intensity average method are significantly increased. The inversion peak errors of 306/974 nm standard bimodal particle system were observed to be 0.170/0.121 and 0.092/0.097, respectively. The peak positions obtained via the iterative recursive method were more accurate, and the experimental results verified the conclusions from the simulation data. The iterative recursive method recalculates the angle weight via successive inversion and comparison of the PSD of each scattering angle. The "correction" effect of angle weight updation can largely offset the PSD error caused by noise, and thus, the "de-noising" performance of resisting the impact of noise has been depicted. Therefore, in a noisy environment, the iterative recursive method should be used for multiangle weighting.

To further improve the accuracy of laser-beam linear datum, a high-accuracy laser collimation system based on a high resolution and high frequency response beam stabilizer and a two-point beam drift separation method was established in this study. First, the translation mirror-based beam stabilizer of the system was studied, and its beam deflection principle and influence factor were analyzed in addition with the two-point beam drift separation method. Then, the resolution and deflection ranges of beam stabilizer, the nonlinearity and hysteresis of piezo, and the associated frequency response were tested. Finally, experimental tests were conducted to determine the accuracy of the laser collimation system. Experimental results indicate that a resolution of 5 nrad of the beam stabilizer can be achieved in deflection range of 120 μrad with frequency response higher than 2 kHz. Furthermore, accuracies of 1.9×10-8 rad and 2.1×10-8 rad of the laser-beam collimation system in 2D directions are obtained, which are both 3 times higher than those corresponding to existing technology. The investigated system is, therefore, capable of satisfying the high accuracy requirements of linear datum using laser-beam.

The accurate calibration of the phase-delay voltage curve is the key to achieving high precision polarization measurements using a nematic Liquid Crystal Phase Variable Retarder (LCVR). An accurate and efficient automatic measurement system was established to improve the accuracy of the liquid crystal retarder delay measurement. Firstly, a new measurement method was proposed, which combined the light intensity method, the Soleil compensator method, and the equal deviation measurement technology, to solve the problem of low measurement accuracy or the low efficiency of the existing methods. A measurement system was developed to overcome these challenges, and automated measurement of the system was realized by Labview technology, which further shortened the system measurement time. Finally, the experimental errors, repeatability, and the working efficiency of the system were verified experimentally. The experimental results show that the system delay measurement error is less than 0.0575%λ, the repeat accuracy is less than 0.0197%λ, and 100 delayed sampling points can be automatically measured in 30 minutes. The system is suitable for the accurate calibration of the phase delay-voltage curve of a liquid crystal variable retarder in the visible range.

A high-precision indoor positioning method based on visible light communication and binocular vision measurement is proposed, with the aims of achieving high precision, low cost, and less interference for indoor mobile robots. This method involves the utilization of a binocular vision sensor to perform imaging measurement of a light-emitting diode(LED)light source.In addition, the three-dimensional (3D) attitude angle of the imaging measurement is recorded using an inertial measurement unit sensor, and the coordinate information of the LED light source is acquired through visible light communication technology. Finally, the 3D coordinates of the binocular vision sensor relative to the LED light source are calculated accurately. A positioning module based on the visible light communication and imaging is developed using this method, in which one or two LED light sources can be used for indoor positioning. When a 1 280×720 resolution image is adopted in the position module, centimeter-level mobile positioning can be obtained in an indoor environment of 2 m×2 m×3 m, and the positioning frequency exceeds 5 Hz. Meanwhile, when using two LED light sources at a positioning distance of 60 cm, not only can the positioning error in the 3D direction be less than 5 cm, but an orientation correction of less than 1.4° can be realized. This proposed method can be applied to provide centimeter-level positioning and navigation services for indoor mobile robots.

The diffraction characteristics of a phase mask for a 520 nm wavelength Femtosecond laser-written FBG are investigated in this study using rigorous coupled-wave theory (RCWT). The results demonstrate that when the phase mask is a rectangular profile, the groove depth and the duty cycle of phase mask must be within the range of 0.57-0.67 μm and 0.32-0.43, respectively, to achieve the desired 0-order diffraction efficiency of less than 2% and ±1-order diffraction efficiency of more than 35%. Using a fused silica phase mask in a 520 nm wavelength femtosecond laser with a period of 1 067 nm, a ruled area 40 mm×30 mm was fabricated via holographic lithography-ion beam etching. The actually produced phase mask was observed to be a trapezoidal profile with a groove depth of 0.665 μm, whose influence on the diffraction efficiency was analyzed. Experimental measurements demonstrated that the 0-order diffraction efficiency was less than 2% and ±1-order diffraction efficiency was more than 40%, which meet the requirement for fabrication of fiber Bragg grating by femtosecond laser.

In a three-dimensional reconstruction task, the pitch range of the multi-layer light detection and ranging (LIDAR) is limited, and the data account of the multi-layer LIDAR is large, which increases the computational burden. A novel method through uniting a single-layer LIDAR and the inertial measurement using a global navigation satellite system/inertial navigation system(GNSS/INS) was developed in this study to solve the problems mentioned above. A single-layer LIDAR scanning feedback plane was used to obtain the depth information, and then a quaternion attitude solution was used to obtain the point cloud image scanned by each station. When using the iterative closest point algorithm to achieve the point cloud registration between sites, the matching point cloud data was filtered and updated to improve the calculation efficiency.The experimental results show that the single-layer LIDAR and the GNSS/INS system can increase the computation rate by 76%, while preserving the digital characteristics of the point cloud. Therefore, the space reconstruction technique using the single-layer LIDAR and GNSS/INS has a similar high registration accuracy as the multi-layer LIDAR three-dimensional reconstruction system and effectively reduces the engineering cost.

The chromatic confocal technique has garnered much attention in the field of surface topography measurement due to its high accuracy, resolution, and rapidity of operation. However, the existing chromatic confocal methods are mostly based on single-point measurement, which restricts their measuring efficiencies.To address this shortcoming, parallel chromatic confocal measurement system based on the chromatic confocal technique was proposed and studied in this paper.In this experimental system, DMD was used as an optical beam splitter, a self-developed dispersive tube lens was applied to produce axial dispersion, and a color camera was used as a photoelectric receiving device. A self-developed color conversion algorithm was also used to simultaneously obtain spectral information from multi-confocal-points.Finally, a parallel chromatic confocal measurement experimental platform was constructed, the experimental measurements corresponding to 50 μm high steps and self-made steps were performed, and the surface topography restoration experiments were conducted.The experimental results indicate that the axial measuring range of the system is 300 μm, and the measuring accuracy is able to achieve micron level.Further, as a 3D measurement system, the proposed technique was verified to be capable of reconstructing the surface topography of a coin based on one-shot.

Owing to the increasing demand of atmospheric sounding data with high precision and high temporal/spatial resolution, onboard infrared hyperspectral sounding systems, such as interferometer,are being developed extensively. Because of the inherent complicated optical structure of interferometer, the calibration precision is highly affected by the status of the instrument.The sensitivity characteristics of calibration precision were analyzed in this study, with regard to the status of the instrument, such as inner calibration blackbody (ICB) emissivity, low temperature external calibration blackbody (ECB) emissivity, temperature difference between the inner blackbody and environment, nonlinear coefficient and darkcurrent (DC) voltage.The results show that: The absolute value of calibration radiation deviation is linearly related to the emissivity of ICB/ECB, and has a positive correlation with the temperature difference between the ICB and environment, nonlinear coefficient, and DC voltage. To achieve the 0.1 K radiometric calibration accuracy, it is imperative to increase the emissivity of ICB/ECB to greater than 0.985. This controls the temperature difference between the ICB and environment, setting it at approximately 0.6 K and the nonlinear effect coefficient at less than 0.04. The influence of the radiometric calibration parameters on calibration radiation combined with the estimation of calibration parameters in the ground vacuum experiment enables initeratively obtaining the optimal estimation of the measured and unknown calibration parameters, thereby enhancing calibration accuracy. The findings of this study are highly significant in the design of the parameters of the infrared hyperspectral interferometer. They also aid in the identification and correction of the source of the radiometric calibration error.

Wavelength-tuned phase-shifting interferometry facilitates phase measurement by changing the wavelength of a tunable diode laser. Random variation of the optical power can cause phase errorsduring the phase-shifting process. To address this problem, an optical power real-time feedback control system and synchronous calibration scheme were developed. First, the measurement error caused by random variation of the optical power was analyzed. An optical power control system was then developed by utilizing suitable optoelectronic detection equipment to convert optical signals into electric signals, such that the optical power can be controlled using a PID. The experimental results show that the accuracy of the optical power is ±0.002 mW and the response speed can approach 600 kHz. Finally, the system was evaluated by detecting the optical component surface.When assembling the system, the PV and RMS are improved by 1.53×10-2λ和2.43×10-3λ, respectively. This system has an important role in high-precision optical component detection.

In order to address the shortcomings of the existing space error analysis methods of CNC machine tools, the geometry error compensation model was optimized based on Abbe principle. First, the geometry error compensation model of the three-axis machine tool was deduced, and the precondition for the correct usage of the model was provided. Second, the mechanism of spatial error transfer of three-axis machine tool was analyzed, based on the Abbe principle. The influence of Abbe error on the positioning accuracy of machine tool was determined, and the theoretical calculation formula was provided and verified by conducting experiments on the moving axis of machine tool. Finally, the existing geometric error compensation model of HTM was optimized based on the Abbe principle and Bryan principle.This model was used to fit the diagonal space error, and compared with the actual measured diagonal error of the machine tool. The existing HTM geometric compensation model can compensate for the machine tool space error, altering it from 41.15 μm to 16.37 μm, with a compensation rate of 60.22%. The optimized compensation model can compensate for the machine tool space error, altering it to 5.32 μm, with a compensation rate of 87.07%, which is an increase of 26.85%. The experimental results show that the optimized compensation model is more reasonable, and further improves the accuracy of space error compensation.

To study the influence of different graphene sound-generator structures on thermoacoustic efficiency, a sound pressure analytical model was established. Theoretical and experimental studies were conducted on the thermoacoustic efficiency of a monolayer graphene sound-generator, multilayer graphene sound-generator, and nickel/chromium-based graphene foam sound-generator. Firstly, the principle of a graphene sound-generator was introduced. A periodic temperature change model and a sound pressure analytical model of this generator were then established. Finally, the three aforementioned sound-generators were studied using an experimental method. The experimental results show that the maximum SPL values of the monolayer, multilayer, and the graphene foam sound-generator are 35.19, 20.36 and 33.42 dB, respectively, in the frequency range of 14~25 kHz when 6 V AC is applied for a distance of 6 cm. The maximum theoretical SPL is approximately 37.45 dB. Graphene sound-generators with lower resistance, lower specific heat capacity, and higher thermal conductivity can achieve higher thermoacoustic efficiency and sound pressure.

In order to solve the thermal control problem of multiple optical remote sensing payloads and platform units on asatellite using limited thermal control resources, a design scheme based on active and passivethermal control strategies was presented. First, according to the satellite characteristics, thermal control requirements, and orbital heat flux, the thermal designing overall project was confirmed. Next, detailed thermal design instructions foroptical payloads and important platform instruments were listed, and the temperaturesof satellite subassemblies are calculated by finite element analysis software. Then, a thermal balance experiment on the whole satellite system was carried out to obtain test temperatures and verify the correctness of the thermal design. Finally, the real effect of the thermal design scheme was proved by comparing the temperature data obtained from on-orbit telemetry, thermal analysis, and thermal testing of the satellite. On-orbit telemetry data indicate that the temperature of the main payload camera is controlled from 19.7 ℃ to 20.3 ℃.The temperature of minor optical payloads ranged from -31.2 ℃ to 6.6 ℃,and the temperature of units inside the satellite cabinranged from 9.7 °C to 29.5 ℃. All the temperature results meet the requirements of the thermal control index. The temperature difference between on-orbit telemetry, thermal analysis, and thermal testing is less than ±3 ℃.The results show that the thermal design of the optical remote sensing satellite is correct and feasible, while the thermal analysis and test process are reasonable and credible.

To improve thecorrosion resistance of the 5xxx aluminum alloy surface,Al0.8FeCoNiCrCux(x=0, 0.25, 0.5, 0.75, 1.0) coatings were prepared on the surface of the alloy vialaser deposition.The crystal texture, micro-structure, andcorrosion resistance of the deposition were analyzed via XRD, SEM and electrochemical workstations.The results demonstrate that with the increase in the quantity of Cu, the phase structure of the HEA changes from BCC1 and BCC2 to BCC1, BCC2, and FCC. Furthermore,when x=0, cracks appeared in the Al0.8FeCoNiCrCux coating, and when x = 0.25, the cracks disappeared, and a light structure was observed in the interdendritic region. With the increase inx, the light structure changed from point distribution to continuous growth.The corrosion resistance of the alloy increased with the decrease in the quantity of Cu, the corrosion current density of Al0.8FeCoNiCrCu0.25 high entropy-alloy coating was estimated to be 7.94×10-8 A/cm2, and the form of corrosion was ascertained to bepitting.Additionally, the corrosion current density of Al0.8FeCoNiCrCuhigh entropy-alloy coating was ascertained to be 8.21×10-7 A/cm2. It also exhibited obvious intercrystalline corrosion,but was still superior to the substrate.This demonstrates that Al0.8FeCoNiCrCuxHEA(0.25≤x≤1) can be used as a coating to enhance the corrosion resistance of aluminum,and its high entropy effect can simultaneously inhibit the formation of inter-metallic compounds caused by the dilution of the substrate. This therefore addresses the problem of crack formation in coatings prepared with traditional materials.

To reduce the influence of adhesion on the process reliability and printing speed in constrained projection fast 3D printing processes. A low-cost constrained substrate scheme was presented with porous transparent support plate and PDMS film for inhibition zone realization. It focused on substrate light transmittance and gas permeability and their influence on the inhibition zone formation and adhesion. The quartz glass substrate with a microporous array was prepared using a laser processing method. The light transmittance, oxygen permeability of the substrate, and thickness of the oxygen inhibition zone were measured experimentally. The influence of the microscopic geometrical characteristics of the substrate on the light transmittance, oxygen permeability, and oxygen inhibition thickness was investigated. 3D printing experiments and a preliminary investigation of the adhesion of the cured layer on the substrate under different substrate conditions were conducted. The results indicate that the micro-pores on the substrate affect the light transmittance and oxygen permeability. The transmittance of the prepared micropores with different geometric features is found to be greater than 84%.The thickness of the oxygen inhibition zone can be controlled by adjusting the oxygen concentration, light intensity, and microstructure of the substrate. The oxygen supply conditions of the substrate affect the thickness of the inhibition zone and subsequently the cued layer peeling force. Adopting the prepared constrained substrateenables the formation of the inhibition layer during printing of the parts and significantly reduces peeling force.The average peeling force of the cured layer from the substrate reduces from 26.4 N to 5.4 N under the experimental conditions in this study.

In a Permanent Magnet Linear Synchronous Motor (PMLSM) with segmented stators, when the mover enters and exits the stator, the control performance was degraded due to the detent force, load disturbance, frictional force, parameter perturbations, and coupling area variations. To address this drawback, a switching control method was proposed. First, an improved Sliding Mode Control (SMC) was used to reduce the speed ripple during the complete coupling stage of the mover and the stator.Then, the chattering caused by the SMC was reduced by the Disturbance Observer (DOB). When the mover exits the stator, thefunction relationships between the electromagnetic parameters and the position of the mover were established to compensate for the speed loss caused by the variation of the coupling area in real time. The results from thesimulation and experiments indicate that the speed ripple is decreased to 0.005 m/s and the setting time is less than 0.3 s in the complete coupling stage.The speed loss is reduced to 0.04 m/s when the mover exits the stator, thereby satisfying the requirements for stability and rapidity of the system.

To meet high tracking stability accuracy requirements, a model of a satellite laser communication coarse tracking system based on a Permanent-Magnet Synchronous Motor (PMSM) was established, and the tracking stability error was analyzed. Owing to a larger tracking error and poor dynamic response performance from the traditional Proportional-Integral-Derivative (PID)control strategy, an improved feed-forward compound control strategy based on the coarse tracking system was proposed.The coarse tracking system was analyzed theoretically to improve dynamic performance. The ground experimental results show that, compared to the traditional control strategy, the improved compound control strategy greatly reduces the dynamic tracking error of the system, from 606 to 13 μrad (measured to one standard deviation, 1δ). The rationality and advancement of the improved compound control strategy are also verified by further on-orbit experiments. Overall performance indicators satisfy the extremely high precision requirements of the satellite laser communication terminal. The proposed control strategy is capable of being used in future designs for other high-performance tracking systems.

Aiming at the small number of feature points and high error rate in the traditional Scale Invariant Feature Transform(SIFT) algorithm, an improved SIFT algorithm was improved based on hyperspectral images. First, hyperspectral images were used as the images generated by Gaussian transformation based on the Gaussian pyramid construction in the traditional SIFT algorithmand the characteristics of hyperspectral images with the same macro-characteristics in different wavebands.This considerably increased the number of real significant feature points detected. Second, the traditional SIFT algorithm and several improved methods only construct the feature descriptor through the pixel information in the neighborhood of the target pixeland ignore the position information of the pixel. In this study, the position information of the target pixel was included in the feature descriptor. The pixel information in the neighborhood was first used for coarse matching, and the position information in the feature descriptor was subsequently used for fine matching. The simulation results showed that by limiting the ratio of the suboptimal value, the method of constructing Gaussian pyramid with hyperspectral images significantly increased the number of feature points extracted, and more extreme points in the image could be extracted.Furthermore, the position information of the target pixel was added to the feature descriptor as the judgment basis of the second stage of feature point matching. Consequently, the number of correct matching was at least 59 times that of the original method, which greatly improved the matching performance of the algorithm.

The feature extraction of a fracture surface is inaccurate due to the lack of local fragments and damage of the geometric features of ornamentation.Thus, in this paper, a method was proposed to solve this problem based on SURF feature descriptor and Jaccard distance. First, Canny operator was used to extract the contour lines of the debris edges and model surface, and a multi-scale space was constructed to extract the feature points of the fault surface. Second, to solve the high redundancy in constructing feature descriptors and high delay in Euclidean distance calculation, SURF feature descriptors with low redundancy was constructed, and then compared the similarity of feature points with Jaccard distance to determine the optimal adjacency relationship of fragments. Finally, the parameters of rigid body were calculated by ICP method, and the fragments were assembled accurately. The experimental results showed that the running time of the algorithm increased by 12%-16%, and the splicing error was at most 0.750 mm. Compared to the traditional method, the proposed method can effectively reduce the considerably large splicing gap and penetration caused by the damage of the fragment model,as well as the splicing error, and achieve fragment splicing efficiently.

Under the influence of temperature changes, the phase of the CCD sampling signal of a high-resolution remote sensing camera canvary.This canadversely affect the signal-to-noise ratio and dynamic range of the imageand may even cause the image to be displayedabnormally. To solve this problem of temperature-relatedsampling position drift, an adaptive compensation design was developed for the CCD sampling position of a high-resolution remote sensing camera. First, the initial position of the CCD sampling signal was precisely adjusted, and an adaptive compensation circuit was then designed tocontrol the power consumption, which rendered the temperature of each driver chipto be essentially the same. The advantage of this method is that the CCD and sampling signals can be effectively monitored during temperature changes, thus, ensuring the accuracy of the sampling position of the CCD signal and stability of the signal-to-noise ratio of the image. The experiments indicate that with this method, the accuracy of the initial position adjustment of the correlated double-sampling signal improves to less than 0.039 ns. Additionally, the maximum delay of the correlated double-sampling signal in the satellite in-orbit temperature range is 0.46 ns, which ensures high-quality imaging and meets the requirements of space applications.

In complex road conditions such as urban tunnels and remote mountainous areas, where satellite navigation system signals arepartially or totally occluded, the accuracy of a vehicle's GNSS/INS integrated navigation system was degraded.Thus, an odometer-assisted high-precision integrated navigation method was proposed in this paper. The combined filtering mode in the method could realize adaptive switching between the GNSS/INS and DR/INS combined modes according to the changes in the vehicle's environment.Additionally, in the combined navigation method, the three-dimensional reckoning position error was extended to the conventional combined navigation filter state. Further, the scale factor error and installation angle error of the odometer could be accurately obtained offline using the odometer error calibration method. However, it is necessary to bind the odometer error parameter to the integrated navigation system in a subsequent use. The vehicle tests reveal that the maximum position error of the integrated navigation system in a single direction of a 7-km signal occlusion scene is 8 m, and the position error is less than 3 m throughout the test.Therefore,the vehicle GNSS/INS integrated navigation system can provide high-precision positioningin complex road conditions.

The Tiny YOLOV3 target detection algorithm has a high error rate for small targets, such as pedestrians, in real-time detection.Therefore, this study aimed to improve the feature extraction network, prediction network, and loss function of the algorithm.First, a two-step convolution layer was added to the feature extraction network to replace the maximum pooling layer in the original network for downsampling. Second, the traditional convolution was replaced with an anti-residual block constructed by a deep convolutional convolution to reduce the model size as well as number of parameters and increase the high-dimensional feature extraction.Third, based on the original two-scale prediction of the network, a scale was added to form a three-scale prediction. Finally, the boundary box position error in the loss function was optimized.The experimental results demonstrat that the improved Tiny YOLOV3 algorithm achieve a target detection accuracy that IS 9.8% higher than the original algorithm, satisfied the real-time requirement, and demonstratedrobustness.The proposed method can better extract target features, and the multi-scale prediction and improvement of the boundary box position error can detect targets more accurately.