A kind of third-order convergence method for solving roots of nonlinear equation is developed by using Newton axiom.It is proved that third-order convergence nears simple root and the computational efficiency of the method is higher than that of other iterative methods.In the case of multiplicity m of roots of equation being known or unknown,the improved formula of the method and the order of convergence are given.Numerical tests are given and the results are compared with those of other iterative methods.It is shown that the proposed method is effective and it has some values in both theory and application.
Emerging artificial intelligence brings new opportunities for embedded machine health monitoring systems. However, previous work mainly focus on algorithm improvement and ignore the software-hardware co-design. This paper proposes a CNN-RNN algorithm for remaining useful life (RUL) prediction, with hardware optimization for practical deployment. The CNN-RNN algorithm combines the feature extraction ability of CNN and the sequential processing ability of RNN, which shows 23%-53% improvement on the CMAPSS dataset. This algorithm also considers hardware implementation overhead and an FPGA based accelerator is developed. The accelerator adopts kernel-optimized design to utilize data reuse and reduce memory accesses. It enables real-time response and 5.89GOPs/W energy efficiency within small size and cost overhead. The FPGA implementation shows 15× CNN speedup and 9× overall speedup compared with the embedded processor Cortex-A9.
The Yellow River is the main source of water for urban and rural area and agricultural irrigation in northern China. Herein, the distribution and risk assessment of perfluoroalkyl acids (PFAAs) were investigated from the Yellow River in Shandong Province, China. The total concentration of PFAAs (∑PFAAs) in surface water and sediments were 37.5–2128 ng/L (mean: 167 ng/L) and not detected−6.95 ng/g dry weight (dw) (mean: 1.02 ng/g dw), respectively. Short-chain PFAAs-perfluorobutanoic acid (PFBA), perfluorohexanoic acid (PFHxA), and perfluorobutane sulfonic acid (PFBS) were the most prevalent PFAAs in surface water. Source analysis showed that firefighting foam (proportion: 31.3 %) and textile treatments and food packaging (proportion: 30.3 %) were the main sources of PFAAs in water. Based on the concentration of PFAAs in water, ecological and potential human health risks were assessed. Perfluorooctanoic acid (PFOA), perfluorononanoate (PFNA), perfluorodecanoate (PFDA) and perfluoroundecanoic acid (PFUnDA) posed nonnegligible ecological risk for some aquatic organisms. Levels of PFAAs (e.g., PFOA, PFNA, and PFDA etc.) in some water samples were higher than the advisory guidelines of PFAAs concentrations in water worldwide, indicating a potential human health risk. Therefore, PFOA, PFNA, PFDA, and PFUnDA are the key focus of pollutants in the water of the Yellow River in Shandong Province, and the standards and limits of these PFAAs in environments including surface water and sediment should be promoted.
Geometry processing in CAD proposes rigorous requirements on mesh quality. In this paper an integrated triangular mesh optimization method is proposed. Edge marking function and local edge operations are used to improve the vertex sampling. Modified weighted centroidal Voronoi tessellation is employed to regularize the triangle geometry. A simulated annealing algorithm is proposed for optimizing the vertex connectivity. Finally, a signal processing filter is developed for mesh denoising. In every optimization stage, the shape deviations from original mesh are prevented, and the boundaries and features are well preserved. Since all modifications are performed locally, the error-prone global parameterization is avoided. This technique has been deployed in real-world product design. Its advantage and robustness are verified by many examples implemented in the geometric design software PUM 2.0 developed by Peking University.
Taking the flocs from cadmium pollution emergency treatment of Longjiang River in Guangxi province as the research object, the stability of the flocs in the simulated static reservoirs and acidic floods was investigated based on the effects of disturbance and pH on the stability of the flocs. The results indicated that disturbance and pH had great effects on the stability of the flocs, and the concentrations of Cd2+ followed the order of pH 5.0 >> pH 6.0 > pH 7. 0 approximately pH 8.0 > pH 9.0 with the original pH of water. When the original pH of water was 5.0, the concentrations of Cd2+ in samples were 19-58 times higher than the national standard limit, and when the original pH of water were 6.0, 7.0, 8.0 and 9.0, respectively, the concentrations of Cd2+ in samples varied from below to 11 times higher than the national standard limit. The release of cadmium from the flocs was higher in the disturbed water, with the concentrations of Cd2+ in most samples higher than 5.0 microg x L(-1), and the highest was double of the national standard limit. In contrast, there was little release in the simulated static reservoirs, with the concentrations of Cd2+ in all samples below 5.0 microg x L(-1), which was lower than the national standard limit. Therefore, the flocs had good stability in the simulated static reservoirs. But it had poor stability in the simulated acidic floods, with higher release of cadmium, and the concentrations of Cd2+ in samples were 14-25 times higher than the national standard limit. Therefore, the monitoring of cadmium concentrations in the floods should be strengthened in the post project analysis for eco-environmental impact of Longjiang River.
Generalized strictly diagonally dominant matrices and nonsingular M-matrices are two kinds of important matrices. In this paper the necessary and sufficient (NS) condition of real square matrix(RSM) to be a generalized strictly diagonally dominant matrix and the NS condition of the RSM′s comparison matrix to be a nonsingular M-matrix are given.A simple and practicable method of judging generalized strictly diagonally dominant matrices and nonsingular M-matrices is introduced. When this method is adopted,only a non-homogeneous linear equation is needed to solve.
The cavity diameter is an important parameter to denote the characteristic of porous concrete. So far, the information about the measurement of cavity diameter of porous concrete is quite insufficient. Especially, the direct measuring method was seldom reported. In the study, a new method to measure the cavity diameter by applying Integrate Charge-Coupled Detector (CCD) imaging technology was developed. The method was simple, time and money spending less and effective. In terms of the method, the mean cavity diameters of porous concretes made by coarse aggregates of 5-10mm,10-16mm, 16-20mm, 20-25mm and 25-32mm were 3.20mm, 4.33 mm, 6.70 mm, 7.75 mm and 9.65mm, respectively. Both void ratio and permeability coefficient of pore concrete were affected by cavity diameter.
We study the reinforcement learning problem of complex action control in the Multi-player Online Battle Arena (MOBA) 1v1 games. This problem involves far more complicated state and action spaces than those of traditional 1v1 games, such as Go and Atari series, which makes it very difficult to search any policies with human-level performance. In this paper, we present a deep reinforcement learning framework to tackle this problem from the perspectives of both system and algorithm. Our system is of low coupling and high scalability, which enables efficient explorations at large scale. Our algorithm includes several novel strategies, including control dependency decoupling, action mask, target attention, and dual-clip PPO, with which our proposed actor-critic network can be effectively trained in our system. Tested on the MOBA game Honor of Kings, our AI agent, called Tencent Solo, can defeat top professional human players in full 1v1 games.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)