This PDF file contains the front matter associated with SPIE Proceedings Volume 13520, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

This work focuses on the highly sensitive and rapid detection of target molecules from small samples through surface-enhanced Raman scattering (SERS). By creating patterned surfaces with adhesion differences using femtosecond laser, we successfully achieved rapid localization of target droplets. The coupling enhancement between the Photonic Nanojets (PNJs) generated by dielectric microspheres and the Localized Surface Plasmon Resonances (LSPRs) generated by metallic nanostructures provides a synergistic enhancement mode produced by hybrid multi-scale structures for high-sensitivity Raman detection. Moreover, to overcome the prolonged evaporation time of droplets on superhydrophobic surfaces, we utilized the Leidenfrost effect to aid evaporation. Compared to natural evaporation, this strategy can overcome diffusion limitations in SERS measurements, allowing diluted analyte molecules to rapidly concentrate within a very short time, and significantly reducing the deposition area of the analyte. The new method of SERS sensing, which involves enriching analytes through heating-evaporation on superhydrophobic/hydrophilic (SH/SHL) substrates and combining them with microspheres, demonstrates a rapid and ultrasensitive biochemical detection strategy, potentially paving the way for future automation in Raman detection.

Microsphere-assisted microscopy has garnered significant attention and rapid development over the past decade due to its characteristics and advantages such as label-free, high resolution, low cost, real-time observation, and ease of integration. Introducing immersion liquid in microsphere-assisted super-resolution imaging not only effectively enhances imaging performance, including contrast, resolution, magnification, and clarity, but also greatly facilitates the adaptation of microsphere-assisted microscopy for applications in biological imaging scenarios. This work proposes a simple and convenient self-assembly method to generate polymer droplet microspheres immersed in liquid with random sizes and positions. The polymer droplet microspheres at random positions primarily have a size distribution in the range of approximately 3-10 μm, with the highest quantity of microspheres having a diameter of around 5 μm. Using such liquid-immersed polymer droplet microspheres as an aid to conventional white-light microscopy enables the observation of periodic grating structures with the minimum feature size of 100 nm on a Blu-ray disc. This study elucidates and explains achieving super-resolution imaging with liquid-immersed polymer droplet microspheres by integrating resolution criteria-based determination and finite element method-based numerical simulation. The research presented here holds promise as a potential candidate for assisting super-resolution imaging using liquid-immersed polymer droplet microspheres.

In the study of multi-pulse laser damage to solar cells, triple-junction GaAs solar cells were irradiated with a pulsed laser of 1.58ms pulse width and 940 nm wavelength to investigate the effects of laser ablation on the surface morphology and performance of the cells. By adjusting the laser irradiation time (one second, three seconds, and five seconds) and varying the laser power density, the surface damage characteristics and changes in current-voltage properties after multi-pulse laser irradiation were systematically analyzed. The experimental results showed that as the laser irradiation time increased, the degree of damage to the solar cells intensified, and their damage threshold gradually decreased. Specifically, the cell surface exhibited varying degrees of melting pits and ablation of the rear electrode. From the perspective of damage mechanisms, the damage to the solar cells caused by the pulsed laser was primarily driven by thermal effects, including thermal melting damage and thermal stress damage. Compared to thermal melting, thermal stress damage had a more significant impact on the electrical performance of the solar cells.

Multi-passed probe light can enhance the signal of atomic magnetometers by increasing the effective optical path length between the alkali metal atomic ensemble and the probe light, thereby strengthening the sensitivity. In recent years, multi-passed probe light has become an important method for improving the response signal and sensitivity of atomic magnetometers. To achieve biaxial magnetic field measurement while maintaining high sensitivity, a biaxial measurement spin-exchange relaxation-free (SERF) atomic magnetometer with a multi-passed probe light is proposed in this paper. The theoretical model and experimental advantages of using multi-passed probe light in longitudinal magnetic field modulation (Z-mode) SERF atomic magnetometers are analyzed. In this paper, a new model of output response of the Z-mode SERF atomic magnetometer based on multi-passed probe light is established, a compact reflective optical path is designed to increase the optical path between the probe light and the polarized alkali metal atomic ensemble, enhance the signal response, and achieve the suppression of low-frequency 1/f noise through longitudinal magnetic field modulation. Experimental results confirmed that the magnetic field sensitivity in the x- and y-axis of the multi-passed probe light configuration improved by 29% and 18%, respectively, compared to the single-pass configuration.

For the landing scenario of spacecraft return, the vehicle optoelectronic system needs to capture the target at the first-time window and perform stable tracking and monitoring over a long period. In this paper, we propose a method for optoelectronic positioning and guidance tracking. This method incorporates collaborative guidance, adaptive detection and tracking, multi-mode positioning, and autonomous prediction of the landing point process. Through algorithm research and the efficient utilization of high-performance computing and network resources for real-time data transmission, we establish a real-time closed-loop control system in conjunction with the optoelectronic system. Field experiments have demonstrated that the proposed method can achieve comprehensive spacecraft monitoring, stable tracking, and precise positioning from a landing site located hundreds of kilometers away, all with a high level of reliability, practicality, and versatility.

With the advancement of laser technology, the development of laser weapons for the destruction of hypersonic vehicles has emerged as a significant area of research. Carbon/silicon carbide (C/SiC) composites are increasingly utilized in the thermal protection systems of hypersonic vehicles due to their outstanding ablation resistance and high-temperature stability. Understanding the thermal damage effects of laser irradiation on these vehicles presents a critical challenge in the field of laser-based missile defense. This study employs numerical simulation to investigate the thermal ablation behavior of thermal protection materials for hypersonic vehicles when subjected to laser irradiation. A two-dimensional ablation model is established, and the unsteady state heat transfer differential equation is solved to obtain the time- dependent temperature field distribution and surface recession of the composite material. The pyrolysis of the material involves phase transition phenomena, which are represented through the relationship between specific heat capacity and thermal conductivity as a function of temperature. Additionally, the dimensionless mass ablation rate is derived at various temperatures, taking into account thermochemical ablation processes, including carbon oxidation, silicon carbide oxidation, and sublimation of C/SiC. COMSOL simulation software is utilized to conduct the thermal ablation simulations of laser-irradiated C/SiC composites. The results demonstrate good agreement with experimental data, thereby validating the accuracy of the proposed computational model.

A mathematical model of the thermal degradation of semiconductor infrared optical materials under continuous near-infrared laser irradiation has been established, which contains the physical processes of temperature field changes and transmittance changes during the laser heating of the materials. The main physical quantities related to the model for three materials, Ge, Si and GaAs are studied comparatively, and the thermal degradation of infrared optical windows made from these three materials under laser irradiation is simulated using the model respectively, and the results are compared. The image blinding effect due to laser-induced thermal degradation of the above three windows is presented using the imaging system simulation software.

This article designs a large aperture multi-band optical window according to design requirements, using a single planar configuration and multispectral materials to ensure its processing performance. The window dimensions are calculated based on empirical formulas, and the stress and strain results of the window under temperature and pressure loads are analyzed through finite element simulation. The window surface shape changes are obtained through polynomial fitting, and the transmitted wavefront is evaluated using optical simulation software. The transmitted wavefront is not greater than λ/ 15. The final design and simulation results meet the requirements of the indicators.

In response to the urgent needs of military optoelectronic systems for field of view range, image quality, and system miniaturization, this article designs a large aperture off-axis tri reflection optical system according to design requirements. The off axis multi reflection structure is used to eliminate system obstruction and reduce structural size. By introducing complex surface shapes, the large aperture design is achieved. While ensuring sufficient resolution, increasing the field of view angle of the optical system can obtain richer target features and improve the system's target detection and recognition capabilities. The design result is an axial MTF ⪆ 0.35@120lp /Mm, RMS wavefront error 0.019 &lambda. The maximum distortion in the full field of view is 0.41%. Sensitivity analysis was conducted based on the actual system installation and debugging process, and the RMS wavefront error was 0.053 after the final system installation and debugging was completed λ. The use of near-infrared and shortwave cameras to image external scenery clearly and with clear details has certain guiding significance for the research of off-axis reflection imaging systems.

Heavy metals in liquid seriously harm the environment and threaten human health. The fluorescence detection system based on fluorescence probe method has the advantages of small size and fast detection speed, which is helpful to understand the pollution status and environmental risk of groundwater and provide scientific basis for liquid environment treatment. However, long time laser irradiation of fluorescent probe produces photobleaching phenomenon, which weakens the fluorescence signal intensity. The excitation noise in the confocal detection optical path is high, resulting in low signal-to-noise ratio, which cannot meet the requirement of high sensitivity detection. In this paper, the anti-interference technology of groundwater heavy metal fluorescence detection device is studied. The effect of photobleaching and optical noise is avoided by single pulse excitation and noise reduction using dichroic mirror. After applying the improvement, the repeatability error is reduced from 1.59% to 0.23%, while the noise is reduced by 87.5%. The stability of the detection results and the signal-to-noise ratio are significantly improved, so the anti-interference ability of detection device is greatly enhanced.

Photoacoustic imaging is a kind of rapidly advancing noninvasive detection technology based on the photon-induced-acoustic effect, which holds the significant potential applications in the biomedical imaging diagnosis field. This technique merges the high contrast capabilities of optical imaging with the deep penetration and high resolution of acoustic imaging. In our work, to study the interaction mechanisms and photoacoustic effects between pulsed laser and capillaries in skin tissue, Finite Element Analysis (FEA) based on COMSOL multiphysis software was employed. Based on the established finite element bio-tissue model, the photoacoustic effect was fully and completely represented as a sequence of light absorption, thermal expansion, and ultrasound wave generation in the bio-tissue, involving the coupling of multiple physical fields and the numerical resolution of partial differential equations. Moreover, in our study, the effects of varying induced wavelengths, capillaries depths in tissue, and absorber sizes on the time-domain photoacoustic signals and photoacoustic intensity were numerical investigated. Numerical results show that the blood in capillaries has stronger photoacoustic effect in the visible light waveband due to the characteristic absorption of hemoglobin in capillaries. The photoacoustic intensity decreases with the increase of capillary radius and depth. To validate the availability of the numerical simulation, the corresponding photoacoustic experiments were performed. In the experiments, we constructed a photoacoustic microscopy detection system and experimentally studied the photoacoustic effects of several influence factors. The experimental results confirm the accuracy of our numerical simulations and provide theoretical insights for future applications of vascular or capillaries photoacoustic imaging in the biomedical field.

This paper aims to develop a hardware system for the precise measurement of the transmittance of optical films. The system mainly consists of a laser light source and a detector, a spectrometer, multi-stage amplification circuits, analog-to-digital conversion circuits, and a power module. Through the optimization of optical path and circuit design, the measurement accuracy and repeatability are ensured. This system is applicable to the performance evaluation of the transmittance of various optical thin film materials and has extensive application prospects.

The advent of services such as big data and cloud computing has driven a continuous increase in the required transmission rate and capacity of optical fiber communication systems, which now support over 90% of global data traffic. Probability Shaping (PS), which adjusts the probability distribution of constellation points, enhances both data transmission rates and the resilience of signals to channel nonlinearity. However, traditional PS methods, such as those using Constant Composition Distribution Matching (CCDM), cannot dynamically adjust the probability distribution of constellation points based on channel conditions. To address this limitation, we propose an end-to-end Autoencoder (AE)-based PS technique for optical communication. Simulation results show that compared to traditional 16/64/256QAM, this scheme achieves a maximum improvement of 0.356 bit/symbol in mutual information over the AWGN channel. In the optical fiber channel, under varying signal power and fiber length conditions, the mutual information is improved in different degrees and the influence of channel nonlinearity on the signal is reduced. Innovatively, residual connections are added to the decoder to enhance the model’s generalization performance, enabling a deeper network structure and improving its ability to handle complex optical fiber channels.

Recording functional signals from various brain regions is crucial in many biomedical studies. For example, monitoring multiple brain regions such as the cortex, the thalamus, and the striatum is necessary under general anesthesia. However, it is still challenging to achieve cortical-wide field-of-view, deep penetrating, high resolution and high recording speed simultaneously. In this study, we proposed the use of epifluorescent microscopy to simultaneously record neuronal activities in the superficial cortex and functional signals from multiple deep brain regions. Neural activity intensity in deep brain regions was relayed to the superficial cortex via optical fibers, with their depths adjusted according to the thickness of cover glass. We have demonstrated the capability to simultaneously record the entire superficial cortical neuronal activities and functional signals of the thalamus and the substantia nigra (SN) during the anesthesia and recovery processes. This method can be integrated with other advanced imaging techniques such as miniscopes and may prove useful in many other applications.

Compared to conventional uniform fibers, long tapered Yb-doped fibers (T-YDF) have outstanding potential to alleviate the detrimental nonlinear effects due to their unique mode field characteristics along the length direction. In this work, a T-YDF-based short-wavelength narrow linewidth monolithic fiber amplifier was constructed through one-stage MOPA configuration. The fiber amplifier emits at the central wavelength of ~1050 nm, and the output lasing performance is carefully studied for different pumping scheme. The experimental results indicated that compared with the co-pumping scheme, the counter-pump one holds a better property involving spectral broadening suppression, SRS suppression and TMI threshold enhancement in the 1050 nm narrow-linewidth fiber lasers. As a result, an output power of 3.56 kW was achieved with a slope efficiency of ~83.6% by adopting bidirectional pumping scheme. At the maximum lasing power, the 3dB spectral linewidth was measured ~0.58 nm, and the SRS suppression ratio was greater than ~27.3dB. This work could offer a promising approach for extending the operating wavelength of high-power narrow linewidth fiber lasers.

The xenon lamp pumping results in an uneven temperature distribution within neodymium glass. This uneven temperature distribution may induce stress birefringence, leading to thermal depolarization and heightened losses in the laser. Identifying the temperature and stress distributions within the material is crucial for accurately predicting the characteristics of beam depolarization. By employing large-diameter neodymium glass as the gain medium and applying the temperature load of the medium in finite element analysis software, the internal stress distribution can be determined. The depolarization distribution of the medium can be calculated theoretically. Although the depolarization distribution of the medium can be calculated theoretically, the thermal depolarization at the four corners of the beam is considerable, with a maximum depolarization of approximately 0.36%. A comparison of the thermal depolarization of neodymium glass across various diameters indicates that the medium diameter is not related to thermal depolarization.

The structure of lepidopteran wing scales plays an important role in the regulation of their biological behaviors, making the 3D imaging of them very useful. However, most of the current three-dimensional microscopic methods are usually destructive to the samples. Fourier light-field microscopy has the characteristics of real-time imaging and multi-angle visualization, thus becomes a preferred method to study the three-dimensional structure of structural color scales in lepidopteran insects. In this study, Fourier light-field microscopy was used to compare the imaging of Morpho helenor and Chrysiridia rhipheus under 10X and 50X objective lenses, where the results demonstrated the difference structure of the wing scale between these two species and the corresponding view angle dependent color variations. This study demonstrates the unique advantages and application prospects of Fourier light-field microscopy for 3D imaging of the large-field thick samples

Herein, we propose a symmetrical double-layer symmetric grating based on lithium niobate with bound states in the continuum (BIC) for tunable dual-band refractive index sensing. By adjusting the structural period and cavity length of the double-layer grating, Symmetry-Protected bound state in the continuum (SP-BIC) and Fabry-Pérot bound state in the continuum (FP-BIC) can be generated in two bands, respectively. Among them, the spacing between the dual dielectric gratings determines the SP-BIC, whereas the FP-BIC is governed through the manipulation of the structural cavity dimensions. Additionally, both forms of BIC can be transformed into quasi-BIC (Q-BIC) states through the alteration of the gap between neighboring gratings or by adjusting the separation between the upper and lower gratings. The simulation results indicate that the dual-band high-performance sensor exhibits a peak sensitivity of 169.5 nm/RIU and a maximum Figure of Merit (FOM) of 157833. Owing to the high electro-optic coefficient of lithium niobate, it shows significant promise for achieving tunable BIC with very low voltage. Subsequently, we showcase the capability to modulate the wavelength of BIC by applying a voltage as low as 1.91μV across the planar electrode arrangements of the lithium niobate grating, achieving a tuning range of approximately 1 nm. Such dual-band high-Q index sensing offers extensive application prospects in the biological and chemical fields.

In recent years, addressing chromatic aberration in metalens has drawn considerable attention. However, ensuring high imaging quality while correcting chromatic aberration remains a significant challenge. In this work, we designed a wide-field achromatic metalens with near-diffraction-limited performance, operating at wavelengths of 480nm, 532nm, and 632nm, and achieving close-to-diffraction-limited Modulation Transfer Function (MTF) at each wavelength within a 40° Field of View (FOV). Our design method innovatively combines phase compensation with a Particle Swarm Optimization (PSO) approach, First, we roughly design the phase dispersion of the metalens phase surface using the compensation phase and then fine-tune the dispersion requirements of the phase surface at different wavelengths using PSO to match our existing unit cell library. Through this method, we achieved excellent chromatic aberration correction using only traditional rectangular unit structures.

This study designs and implements a system for the measurement and analysis of the optical properties of Gaussian thin films, aimed at accurately measuring key optical parameters such as laser reflectance and transmittance. The system employs dual-path data acquisition technology, utilizing laser scanning of the thin film in conjunction with a data acquisition card to achieve efficient data collection and analysis. The software is developed based on the QT framework, integrating OpenGL for the visualization of 3D intensity distribution maps. The system features real-time display, ease of operation, and high-precision measurement, with a final measurement accuracy of up to 1%. Experimental results show that the system can effectively evaluate the optical properties of Gaussian thin films, providing reliable technical support for their measurement and analysis.

Mode Division Multiplexing (MDM) optical fiber communication technology enhances bandwidth utilization and communication capacity to address the increasing demand for high-capacity data transmission in the era of rapid network traffic growth. For MDM optical fiber communication systems affected by various impairments, Optical Performance Monitoring (OPM) can effectively characterize the impairments in received optical signals, directly reflecting the transmission signal state and channel conditions, and help ensure efficient and stable network operation. In this paper, we propose a Multitask Convolutional Neural Network (MT-CNN) aided OPM method for MDM signals. The proposed MT- CNN combines convolutional neural networks and a multitask learning framework to extract the transmission link information and predict the Optical Signal-to-Noise Ratio (OSNR) of few-mode channels synchronously. Simulation results show that the proposed MT-CNN-based OPM scheme can achieve accurate OSNR performance evaluation for MDM fiber transmission signals.

An optical frequency comb-assisted Amplitude Modulated Continuous Wave (AMCW) LiDAR, which features high-precision and fast-speed ranging is proposed. The AMCW signal with a large frequency aperture is downconverted by a single Optical Frequency Comb (OFC), achieving a reduced data volume. In the proof-of-concept experiments, the ranging precision reaches ±0.4 μm while the measurement refresh rate reaches 150 Hz. By utilizing a Micro-Electro-Mechanical System (MEMS) scanning mirror, high-precision three-dimensional laser imaging was also achieved.

We proposed an active micro-scanning super-resolution imaging method. A beam shifter is placed between the lens and the camera and change the tilt angle of the beam shifter’s glass plate to achieve lateral movement of the light. Tilting the beam shifter at a tiny angle in X and Y directions to realize any geometric scanning scheme. By collecting low-resolution image sequences with different shifts in the same scene, a super-resolution algorithm can retrieve the high-resolution. In this work, we adopted the Adaptive Regularization Parameter Tuning (APRT) method to reconstruct the object details. We design and build a prototype of the beam-shifter-based imaging system, and the experiment can demonstrate a 1.5 times imaging resolution improvement of the algorithm.

Multi-aperture system-based high-resolution panoramic image stitching technology, is an effective method to overcome the limitations of spatial bandwidth in single imaging systems. However, significant computational demands and inadequate real-time capabilities frequently challenge panoramic image stitching. However, due to the fast-moving objects in the scene, the stitching result will exist misalignment and artifacts, which affecting the overall quality of the panoramic image. To solve these challenges, we proposed a dynamic panoramic image stitching technique based on a 5×5 compound eye array. We use Compute Unified Device Architecture (CUDA) technology to boost the processes of image registration, furthermore, this method optimizes the stitching result of multiple frames by introducing the Visual Background Extraction (ViBE) algorithm and an advanced seam line search criterion. This method not only reduces the computational complexity associated with stitching adjacent frames, but also eradicates misalignments and artifacts resulting from moving objects. This approach achieves high-quality, rapid stitching of dynamic panoramic images, providing a billion-pixel panoramic image with a horizontal field of view of more than 90 degrees.

Non-Line-of-Sight (NLOS) imaging aims to reconstruct targets or scenes that are directly invisible. When it comes to the research of time-of-flight-based transient NLOS techniques, due to the 3D nature of transient measurements, existing methods require high computational memory, which is not feasible. In this paper, we provide a lightweight U-Net architecture that performs well with low computational cost requirements when reconstructing a hidden target object. Our method makes use of complex convolutional layers with gate mechanisms and channel attention, enabling efficient extraction of local feature details and global information. Extensive experiments show that our method outperforms current methods on simulated data and achieves comparable performance to current methods on real-world data.

We report experimental and numerical simulations on the generation of Deep Ultraviolet (DUV) dispersive waves in a tapered single-ring anti-resonant hollow-core fiber. Initially, we utilize the multi-pass cell compression technology to shorten the Yb-doped laser pulse duration to 35 fs, followed by driving a short piece of tapered anti-resonance fiber with several microjoules of energy, we achieve tunable DUV pulses in the 202 nm to 210 nm wavelength range under moderate gas pressures of six to seven bar. Both experimental and simulation results demonstrate that the tapering process helps reduce the detrimental influence of resonant bands, which otherwise hinder efficient dispersive wave generation. The experimental setup is straightforward and compact, making it feasible for the development of table-top devices for DUV generation, which could address specific requirements in medical and spectroscopy applications.

We present a low-noise Yb-fiber laser based on a Nonlinear Amplifying Loop Mirror (NALM) and an all-Polarization-Maintaining (PM) fiber configuration. Noise optimization is achieved by using a Chirped Fiber Bragg Grating (CFBG) for dispersion compensation, which results in an integrated noise level of -71.83 dB at near-zero dispersion. Further optimization attempts, replacing the CFBG with chirped mirrors, led to an increase in noise, highlighting the critical roles of spectral alignment and intracavity losses in improving performance.

Resonance-enhanced photoacoustic spectroscopy with thin films is an extension of Quartz-Enhanced Photoacoustic Spectroscopy (QEPAS). This technique introduces a resonant coupling between a medium-enhanced Quartz Tuning Fork (QTF) and an acoustic micro-resonator (AmR). In QEPAS, the QTF resonates with the AmR through air, and the strength of this coupling determines the enhancement effect when the AmR is introduced. In this letter, an elastic parylene film was synthesized and applied to the slit of the AmR, with the QTF contact with the film. The pressure distribution within the AmR for different sizes and the resonant frequency of the parylene film were analyzed using Finite Element Modeling (FEM). With the optimized coupling, the elastic parylene film effectively enhanced the resonance between the QTF and the AmR. Additionally, the film acted as a barrier, separating the QTF from the detection gas, which is crucial for the long-term operation of QEPAS in environments with corrosive and dusty gas.

The MEMS Fabry-Perot cavity length can be tuned by moving the MEMS micro mirror to achieve narrowband filtering of the incident continuous spectrum. However, the mirror surface shape would be changed with the deflection of the MEMS micro mirror, affecting the bandwidth and transmittance of the transmission spectrum. By using finite element modeling and simulation, the variation of the surface shapes of different MEMS mirror structures with the mirror deflection were studied. The stress distribution introduced by the mirror deflection were also characterized and the reason for the variation of the surface shapes were also analyzed. By selecting an appropriate elastic beam mounted on the movable micro mirror, the influence of stress introduced by the mirror deflection on the mirror surface shapes could be reduced. The sensitivity of the spectrum filtering performance to the mirror deflection could be lowered to improve the robustness.

Quartz-Enhanced Photoacoustic Spectroscopy (QEPAS) is a highly sensitive technique for trace gas detection and environmental monitoring, leveraging the piezoelectric properties of a Quartz Tuning Fork (QTF) to convert acoustic signals into electrical signals. Traditional QEPAS configurations often face limitations due to environmental noise interference and stringent requirements on laser beam quality. In this work, we present a novel QEPAS spectrophone design incorporating a conical acoustic resonator positioned perpendicularly to the QTF axis and located on one side of the QTF prongs. The conical resonator features a gradual change in cross-sectional area, enhancing acoustic impedance matching and concentrating acoustic energy at the QTF, thereby amplifying the photoacoustic signal. Finite Element Method (FEM) simulations were conducted to model the acoustic field distribution within the resonator, revealing that the acoustic pressure at the QTF prongs increases with the inner radius of the resonator's second side up to an optimal value. The conical acoustic resonator offers superior performance over traditional cylindrical designs, providing enhanced sensitivity, noise reduction, and greater flexibility in practical applications of QEPAS sensors.

The application of neural network algorithms for the quantification of Volatile Organic Compounds (VOCs) concentrations, derived from infrared absorption spectral data, has been shown to achieve superior accuracy compared to conventional least-squares regression techniques. The current neural network models in use generally have issues with low precision and stability, which affect the consistency and credibility of the inversion results. In this study, in order to resolve the aforementioned challenges, we present the pioneering application of the MiniRocket (minimally random convolutional kernel transform) model for the quantitative analysis of VOCs concentrations utilizing hyperspectral data derived from satellite platforms. This model surpasses traditional neural network methodologies by virtue of its automated feature extraction, enhanced computational efficiency, near-deterministic processing, and superior predictive accuracy. The near-deterministic nature of MiniRocket's transformation process ensures reproducibility, as it guarantees identical outcomes given the same input data across diverse computational settings. We employed a training dataset consisting of 120 infrared hyperspectral data with a spectral resolution of 1 cm^-1 and a spectral range of 2.8 to 14.3 μm. Additionally, we utilized a validation dataset comprising 80 sets of test data with randomly assigned concentrations. Experimental results indicate that the MiniRocket model achieves a mean error of prediction (MES) of 6.2×10^-3 parts per million (ppm) for the estimation of pollutant gas concentrations, with a processing time reduced to 0.02 seconds. These outcomes not only underscore the model's superior predictive accuracy but also highlight its unparalleled computational efficiency when compared to other existing models in the field.