New coronavirus disease (COVID-19) has constituted a global pandemic and has spread to most countries and regions in the world. Through understanding the development trend of confirmed cases in a region, the government can control the pandemic by using the corresponding policies. However, the common traditional mathematical differential equations and population prediction models have limitations for time series population prediction, and even have large estimation errors. To address this issue, we propose an improved method for predicting confirmed cases based on LSTM (Long-Short Term Memory) neural network. This work compares the deviation between the experimental results of the improved LSTM prediction model and the digital prediction models (such as Logistic and Hill equations) with the real data as reference. Furthermore, this work uses the goodness of fitting to evaluate the fitting effect of the improvement. Experiments show that the proposed approach has a smaller prediction deviation and a better fitting effect. Compared with the previous forecasting methods, the contributions of our proposed improvement methods are mainly in the following aspects: 1) we have fully considered the spatiotemporal characteristics of the data, rather than single standardized data. 2) the improved parameter settings and evaluation indicators are more accurate for fitting and forecasting. 3) we consider the impact of the epidemic stage and conduct reasonable data processing for different stage.
Parallel manipulators possess the advantages of being compact structure, high stiffness, stability and high accuracy, so such parallel manipulators have been widely employed in application fields as diverse as parallel kinematic machine, motion simulator platform, medical rehabilitation device, and so on. Due to the complexity of the closed-loop structural system, an accurate dynamic model is very difficult to be derived in the absence of some uncertainties parameters and external disturbances. In order to improve the trajectory tracking accuracy with time-varying and nonlinear parameters, this paper addresses the design and implement of adaptive fuzzy sliding mode control (AFSMC) for a three-degree-of-freedom (DOF) parallel manipulator, where internal force term can be linearly separated into a regression matrix and a parameter vector that contains the estimated errors. Furthermore, fuzzy inference unit is utilized to modify the gain parameters in real-time by using the state feedback from the task space and the adaptive law is performed to update uncertainties in dynamic parameters. The proposed controller is deduced in the sense of Lyapunov theory to guarantee the stability while improving the trajectory tracking performance. Finally, simulation experiment results demonstrate that the proposed control method is insensitive to uncertainties and disturbances and permits to decrease the requirement for the bound of these uncertainties, which validate the effectiveness of the developed control method and exhibit good trajectory tracking performance compared with sliding mode control (SMC) and fuzzy sliding model control (FSMC).
This paper presents a novel redundantly actuated 2RPU-2SPR parallel manipulator that can be employed to form a five-axis hybrid kinematic machine tool for large heterogeneous complex structural component machining in aerospace field. On the contrary to series manipulators, the parallel manipulator has the potential merits of high stiffness, high speed, excellent dynamic performance, and complicated surface processing capability. First, by resorting to the screw theory, the degree of freedom of the proposed parallel manipulator is briefly addressed with general configuration and verified by Grübler-Kutzbach (G-K) criteria as well. Next, the inverse kinematics solution for such manipulator is deduced in detail; however, the forward kinematics is mathematically intractable. To deal with such problem, the forward kinematics is solved by means of three back propagation (BP) neural network optimization strategies. The remarkable simulation results of the parallel manipulator demonstrate that the BP neural network with position compensation is an appropriate method for solving the forward kinematics because of its various advantages, such as high efficiency and high convergence ratios. Simultaneously, workspaces, including joint space and workspace of the proposed parallel manipulator, are graphically depicted based on the previous research, which illustrate that the proposed manipulator is a good candidate for engineering practical application.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.