Particle contamination at the head–disk interface (HDI) is a major concern for the long-term reliability of hard disk drives. In the current study, it is shown that, surprisingly, soft and lubricious talc particles have a significant damaging effect on the friction performance of media lubricants, nanometer-thick perfluoropolyethers. Thermogravimetric analysis (TGA) and the adsorption studies indicate that two mechanisms, Lewis acid–catalyzed decomposition and surface adsorption, are responsible for the observed tribological degradation. The potential impacts of the findings here on the design of next-generation media lubricant have been discussed. 相似文献
In the real world, applications with very large state and action spaces and unknown state transition probability, classical reinforcement learning algorithms usually show poor performance. One way to address the performance problem is to approximate the policy or value function. Fuzzy rule-based systems are amongst the well-known function approximators. This paper presents a Flexible Fuzzy Reinforcement Learning algorithm, in which value function is approximated by a fuzzy rule-based system. The proposed algorithm has a separate module for tuning the structure of fuzzy rules. Moreover, the parameters of the system are tuned during the learning phase. Next, the proposed algorithm is applied to the problem of inventory control in supply chains. In this problem, a fuzzy agent (supplier) should determine the amount of orders for each retailer based on their utility for supplier, by considering its limited supply capacity. Finally, a simulation is performed to show the capability of the proposed algorithm. 相似文献
In this paper, a seasonal multi-product multi-period inventory control problem is modeled in which the inventory costs are obtained under inflation and all-unit discount policy. Furthermore, the products are delivered in boxes of known number of items, and in case of shortage, a fraction of demand is considered backorder and a fraction lost sale. Besides, the total storage space and total available budget are limited. The objective is to find the optimal number of boxes of the products in different periods to minimize the total inventory cost (including ordering, holding, shortage, and purchasing costs). Since the integer nonlinear model of the problem is hard to solve using exact methods, a particle swarm optimization (PSO) algorithm is proposed to find a near-optimal solution. Since there is no bench mark available in the literature to justify and validate the results, a genetic algorithm is presented as well. In order to compare the performances of the two algorithms in terms of the fitness function and the required CPU time, they are first tuned using the Taguchi approach, in which a metric called “smaller is better” is used to model the response variable. Then, some numerical examples are provided to demonstrate the application and to validate the results obtained. The results show that, while both algorithms have statistically similar performances, PSO tends to be the better algorithm in almost all problems. 相似文献
Along with the exponential growth of online video creation platforms such as Tik Tok and Instagram, state of the art research involving quick and effective action/gesture recognition remains crucial. This work addresses the challenge of classifying short video clips, using a domain-specific feature design approach, capable of performing significantly well using as little as one training example per action. The method is based on Gunner Farneback’s dense optical flow (GF-OF) estimation strategy, Gaussian mixture models, and information divergence. We first aim to obtain accurate representations of the human movements/actions by clustering the results given by GF-OF using K-means method of vector quantization. We then proceed by representing the result of one instance of each action by a Gaussian mixture model. Furthermore, using Kullback-Leibler divergence (KL-divergence), we attempt to find similarities between the trained actions and the ones in the test videos. Classification is done by matching each test video to the trained action with the highest similarity (a.k.a lowest KL-divergence). We have performed experiments on the KTH and Weizmann Human Action datasets using One-Shot and K-Shot learning approaches, and the results reveal the discriminative nature of our proposed methodology in comparison with state-of-the-art techniques.
In the present paper, the dynamic facilities layout problem is studied in presence of ambiguity of information flow. Product demand (and consequently material flow) is defined as fuzzy numbers with different membership functions. The problem is modeled in fuzzy programming. Three models of expected value, chance-constrained programming and dependent-chance programming and two hybrid intelligent algorithms are then presented. At the end, efficiency of algorithms for solving fuzzy models of dynamic facilities layout is shown through some numerical examples. 相似文献
Analytically based model is presented for behavioral analysis of the plastic deformations in the reinforced materials using the circular (trigonometric) functions. The analytical method is proposed to predict creep behavior of the fibrous composites based on basic and constitutive equations under a tensile axial stress. New insight of the work is to predict some important behaviors of the creeping matrix. In the present model, the prediction of the behaviors is simpler than the available methods. Principal creep strain rate behaviors are very noteworthy for designing the fibrous composites in the creeping composites. Analysis of the mentioned parameter behavior in the reinforced materials is necessary to analyze failure, fracture, and fatigue studies in the creep of the short fiber composites. Shuttles, spaceships, turbine blades and discs, and nozzle guide vanes are commonly subjected to the creep effects. Also, predicting the creep behavior is significant to design the optoelectronic and photonic advanced composites with optical fibers. As a result, the uniform behavior with constant gradient is seen in the principal creep strain rate behavior, and also creep rupture may happen at the fiber end. Finally, good agreements are found through comparing the obtained analytical and FEM results. 相似文献