A novel hybrid modelling framework for GPP estimation: Integrating a multispectral surface reflectance based Vcmax25 simulator into the process-based model.

Terrestrial gross primary productivity (GPP) is the key element in the carbon cycle process. Accurate GPP estimation hinges on the maximum carboxylation rate (Vcmax,025). The high uncertainty in deriving ecosystem-level Vcmax,025 has long hampered efforts toward the performance of the GPP model. Recently studies suggest the strong relationship between spectral reflectance and Vcmax,025. We proposed the multispectral surface reflectance-driven Vcmax,025 simulator using the fully connected deep neural network and built the hybrid modelling framework for GPP estimation by integrating the data-driven Vcmax,025 simulator in the process-based model. The performance of hybrid GPP model was evaluated at 95 flux sites. The result shows that the multispectral surface reflectance-driven Vcmax,025 simulator acquires the satisfactory estimation, with correlation coefficient (R), root mean square error (RMSE) and median absolute percentage error (MdAPE) ranging from 0.34 to 0.80, 14 to 43 µmol m-2 s-1 and 21 % to 66 % across different land cover types, respectively. The hybrid framework generates good GPP estimates with R, RMSE and MdAPE varying from 0.76 to 0.89, 1.79 to 6.16 µmol m-2 s-1 and 27 % to 90 %, respectively. Compared with EVI-driven method, the multispectral surface reflectance significantly improves the Vcmax,025 and GPP estimates, with MdAPE declining by 0.6 %-18 % and 1 % to 21 %, respectively. The Shapley value analysis reveals that red (620-670 nm), near-infrared (841-876 nm) and shortwave infrared (1628-1652 nm and 2105-2155 nm) are the key bands for Vcmax,025 estimation. This study highlights the potential of multispectral surface reflectance for quantifying ecosystem-level Vcmax,025. The new hybrid framework fully extracts the information of all available spectral bands using deep learning to reduce parameter uncertainty while maintains the description of photosynthetic process to ensure its physical reasonability. It can serve as a powerful tool for accurate global GPP estimation.

An FPGA-Implemented Antinoise Fuzzy Recurrent Neural Network for Motion Planning of Redundant Robot Manipulators.

When a robot completes end-effector tasks, internal error noises always exist. To resist internal error noises of robots, a novel fuzzy recurrent neural network (FRNN) is proposed, designed, and implemented on field-programmable gated array (FPGA). The implementation is pipeline-based, which guarantees the order of overall operations. The data processing is based on across-clock domain, which is beneficial for computing units' acceleration. Compared with traditional gradient-based neural networks (NNs) and zeroing neural networks (ZNNs), the proposed FRNN has faster convergence rate and higher correctness. Practical experiments on a 3 degree-of-freedom (DOs) planar robot manipulator show that the proposed fuzzy RNN coprocessor needs 496 lookup table random access memories (LUTRAMs), 205.5 block random access memories (BRAMs), 41 384 lookup tables (LUTs), and 16 743 flip-flops (FFs) of the Xilinx XCZU9EG chip.

Clinical utility of a new immobilization method in image-guided intensity-modulated radiotherapy for breast cancer patients after radical mastectomy.

OBJECTIVE: To investigate clinical utility of a new immobilization method in image-guided intensity-modulated radiotherapy (IMRT) for breast cancer patients after radical mastectomy. MATERIALS AND METHODS: Forty patients with breast cancer who underwent radical mastectomy and postoperative IMRT were prospectively enrolled. The patients were randomly and equally divided into two groups using both a carbon-fiber support board and a hollowed-out cervicothoracic thermoplastic mask (Group A) and using only the board (Group B). An iSCOUT image-guided system was used for acquiring and correcting pretreatment setup errors for each treatment fraction. Initial setup errors and residual errors were obtained by aligning iSCOUT images with digitally reconstructed radiograph (DRR) images generated from planning CT. Totally 600 initial and residual errors were compared and analyzed between two groups, and the planning target volume (PTV) margins before and after the image-guided correction were calculated. RESULTS: The initial setup errors of Group A and Group B were (3.14±3.07), (2.21±1.92), (2.45±1.92) mm and (3.14±2.97), (2.94±3.35), (2.80±2.47) mm in the left-right (LAT), superior-inferior (LONG), anterior-posterior (VERT) directions, respectively. The initial errors in Group A were smaller than those in Group B in the LONG direction (P < 0.05). No significant difference was found in the distribution of three initial error ranges (≤3 mm, 3-5 mm and > 5 mm) in each of the three translational directions for the two groups (P > 0.05). The residual errors of Group A and Group B were (1.74±1.03), (1.62±0.92), (1.66±0.91) mm and (1.70±0.97), (1.68±1.18), (1.58±0.98) mm in the three translational directions, respectively. No significant difference was found in the residual errors between two groups (P > 0.05). With the image-guided correction, PTV margins were reduced from 8.01, 5.44, 5.45 mm to 3.54, 2.99, 2.89 mm in three translational directions of Group A, respectively, and from 8.14, 10.89, 6.29 mm to 2.67, 3.64, 2.74 mm in those of Group B, respectively. CONCLUSION: The use of hollowed-out cervicothoracic thermoplastic masks combined with a carbon-fiber support board showed better inter-fraction immobilization than the single use of the board in reducing longitudinal setup errors for breast cancer patients after radical mastectomy during IMRT treatment course, which has potential to reduce setup errors and improve the pretreatment immobilization accuracy for breast cancer IMRT after radical mastectomy.

A New Finite-Time Circadian Rhythms Learning Network for Solving Nonlinear and Nonconvex Optimization Problems With Periodic Noises.

Nonlinear and nonconvex optimization problems are vital and fundamental problems in science and engineering fields. In this article, a novel finite-time circadian rhythms learning network (called FT-CRLN) is proposed for solving nonlinear and nonconvex optimization problems with periodic noises. Different from the traditional recurrent neural networks, the proposed FT-CRLN can suppress the periodic noise notably and achieve excellent convergence performance in solving nonlinear and nonconvex problems. The theoretical analysis and rigorous mathematical proof verify the superior convergence, high accuracy, and strong robustness of the proposed FT-CRLN. The simulation results demonstrate the effectiveness and robustness of the proposed FT-CRLN in solving nonlinear and nonconvex problems compared with other state-of-art neural networks.