Since the beginning of the 20th century, many researches on the sealing characteristic of mechanical seals were carried out broadly and in depth by various methods and some leakage models were built. But due to the la...Since the beginning of the 20th century, many researches on the sealing characteristic of mechanical seals were carried out broadly and in depth by various methods and some leakage models were built. But due to the lack of the way to characterize the main factors of influence on the leakage, most of the early researches were based on the assumptions that the seal faces topography and the frictional conditions were invariant. In the early built models, the effect of the surface topography change of the seal face on the leakage rate was neglected. Based on the fractal theory, the contact of end faces of the rotary and stationary rings was simplified to be the contact of a rough surface and an ideal rigid smooth surface, and the contact interface＇s cavity size-distribution function as well as the fractal characteristic of the cavity profile curve was discussed. By analyzing the influence of abrasion on the seal face topography and the leakage channel, the time-correlation leakage prediction model of mechanical seals based on the fractal theory was established and the method for predicting the leakage rate of mechanical seals with parallel plane was proposed. The values of the leakage rate predicted theoretically are similar to the measured values of the leakage rate in the model test and in situ test. The experimental results indicate that the leakage rate of mechanical seals is a transient value. The surface topography of the end faces of the seal tings and its change during the frictional wear of mechanical seals can be accurately characterized by the fractal parameters. Under the work conditions of changeless frictional mechanism, the fractal parameters measured or calculated based on the accelerated testing equation can be used to predict the leakage rate of mechanical seal in service. The proposed research provides the basis for determining the leakage state and predicting working life of mechanical seal.展开更多
This paper investigates two distributed accelerated primal-dual neurodynamic approaches over undirected connected graphs for resource allocation problems(RAP)where the objective functions are generally convex.With the...This paper investigates two distributed accelerated primal-dual neurodynamic approaches over undirected connected graphs for resource allocation problems(RAP)where the objective functions are generally convex.With the help of projection operators,a primal-dual framework,and Nesterov's accelerated method,we first design a distributed accelerated primal-dual projection neurodynamic approach(DAPDP),and its convergence rate of the primal-dual gap is O(1/(t^(2)))by selecting appropriate parameters and initial values.Then,when the local closed convex sets are convex inequalities which have no closed-form solutions of their projection operators,we further propose a distributed accelerated penalty primal-dual neurodynamic approach(DAPPD)on the strength of the penalty method,primal-dual framework,and Nesterov's accelerated method.Based on the above analysis,we prove that DAPPD also has a convergence rate O(1/(t^(2)))of the primal-dual gap.Compared with the distributed dynamical approaches based on the classical primal-dual framework,our proposed distributed accelerated neurodynamic approaches have faster convergence rates.Numerical simulations demonstrate that our proposed neurodynamic approaches are feasible and effective.展开更多
Ir-Ta-Ti metal oxide coated titanium anodes prepared by thermal decomposition at different calcination temperatures were characterized by scanning electron microscopy, cyclic voltammetry, and accelerated life test, wh...Ir-Ta-Ti metal oxide coated titanium anodes prepared by thermal decomposition at different calcination temperatures were characterized by scanning electron microscopy, cyclic voltammetry, and accelerated life test, while the consumption rate of an anode was also measured in seawater. The SEM results indicate that all oxide layers exhibit a cracked-mud morphology influenced by calcination temperature. Voltammetric charge (q), obtained by integration of I-E curve, is proportional to the number of surface active sites, and can be taken as a measure of the active surface area of oxide layer. It is found that q* decreases linearly with the increase of calcination temperature, and reaches a maximum at 450℃. The anode prepared at 500℃ possesses the best anodic stability and the longest service life in 1M H2 SO4 at an anodic current density of 2A/cm2. Owing to good corrosion resistance and low consumption rate in seawater, Ir-Ta-Ti metal oxide coated titanium anodes belong to insoluble material with excellent electrocatalytic activity, which are outstanding candidates for using as anode in impressed current cathodic protection systems.展开更多
Gradient descent(GD)algorithm is the widely used optimisation method in training machine learning and deep learning models.In this paper,based on GD,Polyak’s momentum(PM),and Nesterov accelerated gradient(NAG),we giv...Gradient descent(GD)algorithm is the widely used optimisation method in training machine learning and deep learning models.In this paper,based on GD,Polyak’s momentum(PM),and Nesterov accelerated gradient(NAG),we give the convergence of the algorithms from an ini-tial value to the optimal value of an objective function in simple quadratic form.Based on the convergence property of the quadratic function,two sister sequences of NAG’s iteration and par-allel tangent methods in neural networks,the three-step accelerated gradient(TAG)algorithm is proposed,which has three sequences other than two sister sequences.To illustrate the perfor-mance of this algorithm,we compare the proposed algorithm with the three other algorithms in quadratic function,high-dimensional quadratic functions,and nonquadratic function.Then we consider to combine the TAG algorithm to the backpropagation algorithm and the stochastic gradient descent algorithm in deep learning.For conveniently facilitate the proposed algorithms,we rewite the R package‘neuralnet’and extend it to‘supneuralnet’.All kinds of deep learning algorithms in this paper are included in‘supneuralnet’package.Finally,we show our algorithms are superior to other algorithms in four case studies.展开更多
基金supported by China Postdoctoral Science Foundation (Grant No. 20070410323)Jiangsu Provincial Planned Projects for Postdoctoral Research Funds of China (Grant No. 0701001C)Jiangsu Provincial Planned Projects for Fostering Talents of Six Scientific Fields of China (Grant No. 07-D-027)
文摘Since the beginning of the 20th century, many researches on the sealing characteristic of mechanical seals were carried out broadly and in depth by various methods and some leakage models were built. But due to the lack of the way to characterize the main factors of influence on the leakage, most of the early researches were based on the assumptions that the seal faces topography and the frictional conditions were invariant. In the early built models, the effect of the surface topography change of the seal face on the leakage rate was neglected. Based on the fractal theory, the contact of end faces of the rotary and stationary rings was simplified to be the contact of a rough surface and an ideal rigid smooth surface, and the contact interface＇s cavity size-distribution function as well as the fractal characteristic of the cavity profile curve was discussed. By analyzing the influence of abrasion on the seal face topography and the leakage channel, the time-correlation leakage prediction model of mechanical seals based on the fractal theory was established and the method for predicting the leakage rate of mechanical seals with parallel plane was proposed. The values of the leakage rate predicted theoretically are similar to the measured values of the leakage rate in the model test and in situ test. The experimental results indicate that the leakage rate of mechanical seals is a transient value. The surface topography of the end faces of the seal tings and its change during the frictional wear of mechanical seals can be accurately characterized by the fractal parameters. Under the work conditions of changeless frictional mechanism, the fractal parameters measured or calculated based on the accelerated testing equation can be used to predict the leakage rate of mechanical seal in service. The proposed research provides the basis for determining the leakage state and predicting working life of mechanical seal.
基金supported by the National Natural Science Foundation of China (Grant No.62176218)the Fundamental Research Funds for the Central Universities (Grant No.XDJK2020TY003)。
文摘This paper investigates two distributed accelerated primal-dual neurodynamic approaches over undirected connected graphs for resource allocation problems(RAP)where the objective functions are generally convex.With the help of projection operators,a primal-dual framework,and Nesterov's accelerated method,we first design a distributed accelerated primal-dual projection neurodynamic approach(DAPDP),and its convergence rate of the primal-dual gap is O(1/(t^(2)))by selecting appropriate parameters and initial values.Then,when the local closed convex sets are convex inequalities which have no closed-form solutions of their projection operators,we further propose a distributed accelerated penalty primal-dual neurodynamic approach(DAPPD)on the strength of the penalty method,primal-dual framework,and Nesterov's accelerated method.Based on the above analysis,we prove that DAPPD also has a convergence rate O(1/(t^(2)))of the primal-dual gap.Compared with the distributed dynamical approaches based on the classical primal-dual framework,our proposed distributed accelerated neurodynamic approaches have faster convergence rates.Numerical simulations demonstrate that our proposed neurodynamic approaches are feasible and effective.
文摘Ir-Ta-Ti metal oxide coated titanium anodes prepared by thermal decomposition at different calcination temperatures were characterized by scanning electron microscopy, cyclic voltammetry, and accelerated life test, while the consumption rate of an anode was also measured in seawater. The SEM results indicate that all oxide layers exhibit a cracked-mud morphology influenced by calcination temperature. Voltammetric charge (q), obtained by integration of I-E curve, is proportional to the number of surface active sites, and can be taken as a measure of the active surface area of oxide layer. It is found that q* decreases linearly with the increase of calcination temperature, and reaches a maximum at 450℃. The anode prepared at 500℃ possesses the best anodic stability and the longest service life in 1M H2 SO4 at an anodic current density of 2A/cm2. Owing to good corrosion resistance and low consumption rate in seawater, Ir-Ta-Ti metal oxide coated titanium anodes belong to insoluble material with excellent electrocatalytic activity, which are outstanding candidates for using as anode in impressed current cathodic protection systems.
基金This work was supported by National Natural Science Foun-dation of China(11271136,81530086)Program of Shanghai Subject Chief Scientist(14XD1401600)the 111 Project of China(No.B14019).
文摘Gradient descent(GD)algorithm is the widely used optimisation method in training machine learning and deep learning models.In this paper,based on GD,Polyak’s momentum(PM),and Nesterov accelerated gradient(NAG),we give the convergence of the algorithms from an ini-tial value to the optimal value of an objective function in simple quadratic form.Based on the convergence property of the quadratic function,two sister sequences of NAG’s iteration and par-allel tangent methods in neural networks,the three-step accelerated gradient(TAG)algorithm is proposed,which has three sequences other than two sister sequences.To illustrate the perfor-mance of this algorithm,we compare the proposed algorithm with the three other algorithms in quadratic function,high-dimensional quadratic functions,and nonquadratic function.Then we consider to combine the TAG algorithm to the backpropagation algorithm and the stochastic gradient descent algorithm in deep learning.For conveniently facilitate the proposed algorithms,we rewite the R package‘neuralnet’and extend it to‘supneuralnet’.All kinds of deep learning algorithms in this paper are included in‘supneuralnet’package.Finally,we show our algorithms are superior to other algorithms in four case studies.
