Fanghao Lu

Monitoring combustion field parameters in complex environments, particularly within combustion chambers of aerospace engines, holds substantial importance for advancing propulsion technology. However, the high-temperature, high-pressure, and highly dynamic characteristics of combustion parameters, coupled with complex testing environments, present formidable challenges to dynamic parameter monitoring techniques. Laser absorption spectroscopy technology, with its nonintrusive nature and high-temporal-resolution capabilities, is considered a highly promising approach for such demanding applications. This paper provides a comprehensive review and summary of the latest technological advances in laser absorption spectroscopy over the past decade, along with its most recent applications in combustion diagnostics. Furthermore, the paper analyzes potential future development directions for this critical measurement technology in addressing the evolving needs of advanced combustion systems.

At-home rehabilitation for post-stroke patients presents significant challenges, as continuous, personalized care is often limited outside clinical settings. Moreover, the lack of integrated solutions capable of simultaneously monitoring motor recovery and providing intelligent assistance in home environments hampers rehabilitation outcomes. Here, we present a multimodal smart home platform designed for continuous, at-home rehabilitation of post-stroke patients, integrating wearable sensing, ambient monitoring, and adaptive automation. A plantar pressure insole equipped with a machine learning pipeline classifies users into motor recovery stages with up to 94% accuracy, enabling quantitative tracking of walking patterns during daily activities. An optional head-mounted eye-tracking module, together with ambient sensors such as cameras and microphones, supports seamless hands-free control of household devices with an average latency under 1 s with consistent operation. These data streams are fused locally via a hierarchical Internet of Things (IoT) architecture, ensuring low latency and data privacy. An embedded large language model (LLM) agent, Auto-Care, continuously interprets multimodal data to provide real-time interventions—issuing personalized reminders, adjusting environmental conditions, and notifying caregivers. Implemented in a post-stroke context, this integrated smart home platform increased mean user satisfaction from $3.9~\pm ~0.8$ in conventional home environments to $8.4~\pm ~0.6$ with the full system (n = 20). Beyond stroke, the system offers a scalable, patient-centered framework with potential for long-term use in broader neurorehabilitation and aging-in-place applications.

This article presents a novel method for flowrate measurement of gas-liquid two-phase flow using a throat-extended Venturi tube (TEVT) combined with a microwave resonant cavity (MRC). First, a total flowrate model is developed from differential pressure (DP) equations across the convergent and straight sections of the TEVT. Subsequently, a gas-liquid flowrate ratio model is established using microwave resonant frequency and multiple DP fluctuations, fitted via support vector machine. The particle swarm optimization algorithm then solves the coupled equations for gas and liquid flowrates. The proposed method integrates two sensing modalities and leverages static and dynamic features of raw measurement data to achieve simultaneous gas-liquid flowrate prediction. Experiments were conducted on a gas-liquid two-phase flow test facility using the TEVT and MRC. Within gas flowrate of 0.07 ~ 8.09 m³/h and liquid flowrate of 1.12 ~ 13.64 m³/h, the relative errors of the predicted liquid and gas flowrates are within ±5% and ±9.5%, respectively. The results demonstrate that the proposed method performs well in predicting gas-liquid two-phase flowrate, offering potential for extension to broader range of flow conditions.

This article presents a novel method for measuring the mass flow rate of gas-liquid two-phase flow based on the multi-sensor data fusion and soft computing model. A multi-sensor system comprising a throat-extended Venturi tube and a dual-modality electrical sensor has been developed for gas-liquid two-phase flow measurement. Soft computing models are employed to address the intricate non-linear mapping between the measurement data and flow parameters. Initially, flow regimes are identified based on the time-domain features of the multi-sensor data using a support vector machine. Subsequently, mass quality is derived from the multi-differential pressure fluctuations and the eigenvalue sequence of the normalized electrical matrices, employing a hybrid neural network comprising a convolution neural network and a deep neural network. Ultimately, gas/liquid over-reading is predicted via Extreme Gradient Boosting using multi-differential pressure ratios. The gas and liquid mass flow rates are subsequently derived from the preceding results. The proposed method addresses the issue that the parameters measurement of gas-liquid two-phase flow is significantly influenced by the flow regimes, and achieves accurate flow rate measurement under the diverse flow regimes. Experimental validation confirms the method’s effectiveness and superior performance compared to conventional approaches.

Wavelength modulation spectroscopy (WMS) is recognized as a good noise immunity method in tunable diode laser absorption spectroscopy (TDLAS). It can obtain high signal-to-noise ratio harmonic intensity signals for precise gas parameter measurement, like temperature and concentration. However, the TDLAS-based velocity is sensitive to the Doppler shift in the laser frequency caused by the high-speed flow, which usually requires complex wavelength calibration and precise signal alignment. This work proposed a novel method for the precise velocity measurement directly by extracting the time interval of the even-symmetric 2f/1f harmonics signal with a self-convolution algorithm. The even-symmetric 2f/1f harmonics signal is obtained from the up-and-down laser optical frequency scanning WMS with lock-in amplification technology. A portable measurement system has been developed to achieve simultaneous velocity and temperature measurement of the gas in turbojet engine exhaust with different operation states. The maximum standard deviation of the velocity derived by the proposed method is 15.6 m/s, while that by traditional DAS and WMS are 69.2 m/s and 26.4 m/s, respectively. Experiment results have verified the improvement of velocity precision by the proposed method. It can provide reliable experimental data for improving the design of combustors and optimizing numerical combustion simulations.

Tunable diode laser absorption spectroscopy (TDLAS) is a popular technique for non-contact temperature measurement but is highly susceptible to beam-steering distortion due to the time-varying along-path inhomogeneity of refractive index in dynamic process monitoring. An adaptive beam-steering-distortion compensation method is proposed to eliminate the distortion in TDLAS measurement. A high-frequency modulated laser at a non-absorbing wavelength is coupled with a conventional low-frequency scanning laser at an absorbing wavelength and passes through the target field along an identical path. A synchronous demodulation method is proposed to demodulate the transmitted hybrid laser signal into the distorted absorption and distortion signals. The distortion signal is utilized to adaptively compensate the distortion parasitic on the distorted absorption signal. Experimental results obtained on a vibrated water thermostat show that the maximum deviation of temperature derived by the proposed method is 4.97 K, while that by the traditional DAS method is 34.64 K. Experimental results on a turbulent flame show that the relative standard deviation of temperature obtained by the proposed method has decreased from 36.7 % (DAS) to 4.20 %. Also, the temperature fitting success rate has increased from 88.0 % to 100 % in the turbulent flame case. The results show that the proposed method is more precise and robust than the traditional method. The proposed method will potentially play an important role in laser-based dynamic combustion monitoring and diagnosis.

Precise and bias-immune absorption spectrum was directly extracted from wavelength-modulated intensities for dynamic gas temperatures by a proposed novel downsampling scheme. Absorbed wavelength-modulated laser intensity generated multiple direct absorption profiles by downsampling at equal phases in each modulation cycle. The number of achieved profiles equals to that of sampled intensities in each wavelength modulation cycle. All these profiles are employed to remove intensity biases from thermal radiations and obtain an entire absorption spectrum of target gases. A water-based thermostat was used to evaluate the noise reduction of the proposed method in controlled noisy cases. Deviations of achieved gas temperatures are less than 4°C, and precision is quantitated by SNR values that achieved by the proposed method are respectively 15 dB and 5 dB higher than those of direct absorption spectroscopy and wavelength modulation spectroscopy methods. Dynamic temperatures measured by the proposed method are used to evaluate temporal variations of an acoustically excited flame. It identified the sinusoidal temperature variation, and its harmonic strength at the target frequency is 1.5 times stronger than that of the direct absorption spectrum method in noisy cases, e.g. SNR from 40 to 60 dB. The intensity bias from the thermal radiation is also derived by the proposed method, and the bias value agrees well with that of the reference detector for radiative variations of a correlation coefficient of 0.98. Both simulated and experiment results verified the robust performance of the proposed method for flame temperatures with strong thermal radiation and noises.

Laser absorption spectroscopy (LAS) tomography yields cross-sectional images of flame temperature and gas molar concentration in combustions. The total laser absorbance of a specific gas, such as water vapor at each point in the region of interest, was obtained from local normalized second harmonics on limited detectors. These detectors restrict the spatial resolution during image reconstructions through typical iterative methods. A sparse Zernike fitting method was introduced for LAS tomography, as the temperature in real physical reality is smoothly distributed. The distributions of total absorbance were sparsely fit as finite Zernike polynomials, and the coefficients were mostly zero or about zero. The number of coefficients was much smaller than that of pixels, i.e., unknowns in iterative methods. The underdetermined inverse problem was effectively alleviated, and a higher spatial resolution was achieved. A fast convergence and effective method for optimal step selection was also introduced for the iterative solution of the sparse model. The effectiveness of the proposed method was validated by both simulated and experimental data for temperature and concentration imaging. Performance comparisons were made with the typical iterative methods, such as SART and Landweber methods; temporal variations of reconstructed distributions for dynamical flames were captured with higher signal-to-noise ratios. Both the fundamental frequencies of 8.797 Hz without acoustic excitation and 12.7 Hz in the case of 150-Hz excitation were successfully captured, and tomographic images agreed well with the dynamical flame evolutions.

A back propagation (BP) neural network is introduced and implemented on a field programmable gate array (FPGA) chip into tunable diode laser absorption spectroscopy (TDLAS) tomography for fast response and a high signal-to-noise ratio (SNR) in temperature and water concentration imaging. The network implemented on the FPGA extracts the peak values of normalized second harmonics of absorption spectra. A recursive demodulator based on a Kalman filter generates outputs of the training database. Compared with the typical quadrature demodulator, the sampling time of the BP neural network drops to one-quarter of the typical one, with a 5 dB increase in the SNR at different noise levels. To verify and evaluate the proposed method, the FPGA-based 32-bit fixed-point BP neural network was compared with the floating-point BP neural network and the quadrature demodulator in numerical simulations. In the real experiments, Bunsen burner flames with acoustic excitation were imaged, and noisy distortions of the normalized second harmonics and average temperature were effectively reduced by using the proposed method. Temporal fluctuations of up to 750 Hz can be clearly identified from variations of reconstructed distributions of temperature and water vapor molar concentration, in which cases the typical quadrature demodulator failed to work. These results showed that the proposed method works well in the FPGA chip and reveals more details in dynamic flame monitoring.

In this paper, an AC-based agile parallel hardware system was introduced for ECT. A multi-frequency excitation and measurement strategy was proposed to minimize the switching time between different channels. For an ECT sensor with 2 ^m electrodes, all the mutual capacitances can be measured much faster, which is m/(2 ^m - 1) of the time used by the classical single-frequency excitation and measurement strategy. It implies that the proposed strategy accelerates the capacitance acquisition in an exponential way. Besides the efforts made to reduce the switching times between channels, the operating time of each electronic switch was selected to fasten the demodulation. The hardware fabricated here can be modified agilely by running codes, and different excitation and demodulation methods can be implemented on the chip. To verify the performance of the agile system, ECT sensors with different electrodes are facilitated to measure the mutual capacitances. The capacitances can be achieved with a frame rate of respectively up to 15.15 kilo frames per second (kfps) and 12.82 kfps in the cases of 12 and 16 electrodes. The ignition process of a flame generated by a Bunsen burner can be captured by the proposed system. Also, the discharging of the igniter can be imaged, and it is usually undistinguishable in a low-speed system.