High-resolution three-dimensional temperature fields are essential for studying flame combustion, and tunable diode laser absorption tomography (TDLAT) is an effective method for diagnosing flame combustion conditions. In actual combustion measurements, the reliance of TDLAT on line-of-sight (LOS) measurements leads to limited data and reduced dimensionality in analyzing combustion fields. This study proposes a method using area-array sensor-coupled absorption spectroscopy to measure the three-dimensional temperature field of flame accurately, aiming for enhanced combustion diagnosis. The laser beam is configured into a cone shape, and after traversing the combustion field under examination, the area-array sensor receives a projection signal. This signal is then used to reconstruct a high-resolution, multidimensional temperature field. We confirmed the accuracy and robustness of the algorithm through numerical simulations and compared these with experimental results from the TDLAT setup. Our TDLAT detection system demonstrates high precision and effectively measures temperature fields in complex flame imaging scenarios.

The combustion temperature field is a key indicator of combustion stability, equipment integrity, and abnormal operating conditions. This study addresses short-horizon forecasting of laboratory-scale apparent infrared temperature fields using a Convolutional Long Short-Term Memory (ConvLSTM) network. A chronological 70/15/15 split is applied to 1800 available sequences, with the first 10 frames of each sample treated as observed inputs. Proper orthogonal decomposition (POD) is applied exclusively to the input frames at a 95% cumulative energy retention threshold, and predictions are evaluated on subsequent frames. POD is therefore used as an input-side low-rank modal preprocessing method to extract dominant coherent structures and suppress redundant variations, rather than as a reduction in the ConvLSTM network architecture itself. Performance is assessed via normalized domain PSNR/SSIM and Kelvin domain errors. The first three forecast steps maintain appreciable image similarity, while the fourth and fifth steps exhibit marked degradation. Therefore, conclusions are restricted to short-horizon forecasting, with no assertions regarding closed loop combustion control or cross-setup generalization. Simultaneously, ablation experiments with the same architecture, including and excluding the POD preprocessing step, quantified the individual contribution of the POD preprocessing step.