In smart cities and factories, robotic applications require high dexterity and security, which requires precise inverse dynamics model. However, the physical modeling methods cannot model the uncertain factors of the manipulator such as flexibility, joint clearance and friction, etc. As an alternative, artificial intelligence (AI) techniques have become increasingly popular in robotics for smart cities and factories. In this paper, deep learning neural network based on LSTM (Long Short-Term Memory) is adopted to predict the manipulator inverse dynamics. This study aims to summarize the influence of the hyper-parameter settings on model performance and to explore the applicability of the LSTM model to joint torque prediction of multiple degrees of freedom series manipulator. Furthermore, the feasibility of using only joint position as input data for torque prediction is verified. Simulation result has shown that, for the proposed deep learning architecture, the effects of the number of maximum epochs on model performance should be prioritized. The effects of the number of hidden nodes on model performance are limited, while prediction accuracy will deteriorate as the number of hidden layers increases. It is proved that it is feasible to predict inverse dynamics when input data is joint position only. The experimental results show that the training time increases with the increase of hidden layers, neurons and epochs. INDEX TERMS Smart cities and factories, inverse dynamics, robot, green computing, deep learning, LSTM.
In this article, a direct and inverse iteration design method for mixed-flow pumps was introduced, and a three-dimensional design platform for mixed-flow pumps was established. Through iteration calculation of two kinds of stream surfaces to solve both the continuity equation and motion equation of the fluid at the same time, and therefore obtaining the quasi-three-dimensional flow field inside the mixed-flow pump, the accuracy of the flow field calculation was improved. By repeating iteration of the direct calculation and inverse design, influence of blade shape on flow field calculation was fully considered, thus assuring that the final design of blade camber line complies with actual flow pattern. After comparing and analyzing the meridional velocity distribution, as well as the relative circulation distribution and expelling coefficient in the blade zone, it was proved that compared with others designed by conventional methods, the impeller designed with three-dimensional design platform has a better hydraulic performance, especially the blade energy conversion capacity that witnessed a significant improvement. With this three-dimensional design platform, a parametrization was applied to velocity moment distribution and blade leading and trailing edges positions. The influence of velocity moment distribution parameters P and a0, and the blade leading and trailing edges positions parameters θh and θt on blade wrap angle and the internal flow of the impeller were analyzed. By curve fitting, the functional relationship between blade wrap angle ϕ and parameters θt, P and a0 was obtained. The blade wrap angle control strategy during the process of mixed-flow pump impeller design was then put forward, thus realizing three-dimensional design of mixed-flow pump blades with controllable blade wrap angle. The performance test for the mixed-flow pump model suggested that the three-dimensional design method with controllable blade wrap angle, by controlling blade wrap angle size and adopting direct and inverse iteration design method, ensures a better hydraulic and cavitation performance for the mixed-flow pump impeller designed with three-dimensional design platform.
Based on the three-dimensional design platform, this article conducted parametrization of blade leading and trailing edge positions, and analyzed the influence of different position parameters on mixed-flow pump hydraulic performance and internal flow with model test and numerical simulation. The results showed that the hydraulic efficiency of a mixed-flow pump increased slightly when the position parameter h of the blade leading edge on the hub increased, and increased significantly when the position parameter t of the blade leading and trailing edge positions on the shroud increased. However, with t increasing, the growth rate of decreased. Numerical simulation has shown that by selecting a proper value of t , the impeller energy conversion capacity can be effectively improved, and the distributions of static pressure and total energy can be more uniform in the flow passage. Meanwhile, with t increasing, blade angle on the blade trailing edge decreased. Correspondingly, the absolute velocity in the outlet zone decreased, and the hydraulic loss in the outlet zone also decreased, which is beneficial to improving hydraulic efficiency of the mixed-flow pump. Within the value range of 7-9 , with different combinations of position parameter t1 of the blade leading edge on the shroud and position parameter t2 of the blade trailing edge on the shroud, the mixed-flow pump hydraulic efficiency and blade wrap angle ' show a linear positive correlation, suggesting that an increase in ' could significantly improve impeller energy conversion capacity. Compared to t1 , t2 has a more significant influence on ' and . Thus, the value of t2 should be carefully attended to during design process.
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.