Giant magneto-impedance(GMI)effect has been realized in the sandwiched FeSiB/Cu/FeSiB films.With magnetic field Ha and ac current applied along the longitudinal direction of the sample,the GMI ratio increases with the...Giant magneto-impedance(GMI)effect has been realized in the sandwiched FeSiB/Cu/FeSiB films.With magnetic field Ha and ac current applied along the longitudinal direction of the sample,the GMI ratio increases with the increasing Ha,reaching a positive maximum value,and then decreases to negative values with further increase of magnetic field.Field dependence of the GMI ratio also indicates that the magnetic field corresponding to the maximum GMI ratio is different for various frequencies.The positive maximum GMI ratio is 17.2%for Ha=1600A/m and frequency of 3 MHz.In addition,the films display a large negative GMI ratio with a magnetic field applied along the transverse direction and the value of the GMI ratio is about-13.4%for Ha=5600A/m and frequency of 3 MHz.展开更多
Coupled Hidden Markov Model (CHMM) is the extension of traditional HMM, which is mainly used for complex interactive process modeling such as two-hand gestures. However, the problems of finding optimal model parameter...Coupled Hidden Markov Model (CHMM) is the extension of traditional HMM, which is mainly used for complex interactive process modeling such as two-hand gestures. However, the problems of finding optimal model parameter are still of great interest to the researches in this area. This paper proposes a hybrid genetic algorithm (HGA) for the CHMM training. Chaos is used to initialize GA and used as mutation operator. Experiments on Chinese TaiChi gestures show that standard GA (SGA) based CHMM training is superior to Maximum Likelihood (ML) HMM training. HGA approach has the highest recognition rate of 98.0769%, then 96.1538% for SGA. The last one is ML method, only with a recognition rate of 69.2308%.展开更多
A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The ou...A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The output of SVM is moderated to be probability output, which replaces the Mixture of Gauss (MOG) in HMM. Wavelet transformation is used to extract observation vector, which reduces the data dimension and improves the robustness.The hybrid system is compared with pure HMM face recognition method based on ORL face database and Yale face database. Experiments results show that the hybrid method has better performance.展开更多
Both sputtering conditions and crystallizing temperatures have great influence on the microstructures and phase transformation characteristics for Ti 51 Ni 44Cu 5. By means of the resistance temperature measurement, X...Both sputtering conditions and crystallizing temperatures have great influence on the microstructures and phase transformation characteristics for Ti 51 Ni 44Cu 5. By means of the resistance temperature measurement, X ray diffraction and atomic fore microscopic study, the results indicate that the transformation temperatures of the thin films increase and the "rock candy" martensitic relief is more easily obtained with promoting the sputtering Ar pressure, sputtering power, or crystallizing temperature. However, when sputtering Ar pressure, sputtering power, or crystallizing temperature are lower, a kind of "chrysanthemum" relief, which is related with Ti rich GP zones, is much easier to be observed. The reason is that during crystallization process, both of the inherent compressive stresses introduced under the condition of higher sputtering pressure or higher crystallizing temperature are helpful to the transition from GP zones to Ti 2(NiCu) precipitates and the increase of the transformation temperatures. The addition of copper to substitute for 5% nickel in mole fraction can reduce the transformation hysteresis width to about 10 ～ 15?℃.展开更多
The application of reinforcement learning is widely used by multi-agent systems in recent years. An agent uses a multi-agent system to cooperate with other agents to accomplish the given task, and one agent′s behavio...The application of reinforcement learning is widely used by multi-agent systems in recent years. An agent uses a multi-agent system to cooperate with other agents to accomplish the given task, and one agent′s behavior usually affects the others′ behaviors. In traditional reinforcement learning, one agent takes the others location, so it is difficult to consider the others′ behavior, which decreases the learning efficiency. This paper proposes multi-agent reinforcement learning with cooperation based on eligibility traces, i.e. one agent estimates the other agent′s behavior with the other agent′s eligibility traces. The results of this simulation prove the validity of the proposed learning method.展开更多
This paper presents temporal memory reinforcement learning for the autonomous micro-mobile robot based-behavior. Human being has a memory oblivion process, i.e. the earlier to memorize, the earlier to forget, only the...This paper presents temporal memory reinforcement learning for the autonomous micro-mobile robot based-behavior. Human being has a memory oblivion process, i.e. the earlier to memorize, the earlier to forget, only the repeated thing can be remembered firmly. Enlightening forms this, and the robot need not memorize all the past states, at the same time economizes the EMS memory space, which is not enough in the MPU of our AMRobot. The proposed algorithm is an extension of the Q-learning, which is an incremental reinforcement learning method. The results of simulation have shown that the algorithm is valid.展开更多
A novel MEMS variable optical attenuator (VOA), which has completely different attenuation mechanism from those in literatures, is proposed and demonstrated in this paper. The basic operation principle is that the opt...A novel MEMS variable optical attenuator (VOA), which has completely different attenuation mechanism from those in literatures, is proposed and demonstrated in this paper. The basic operation principle is that the optical power coupled between two initially aligned single-mode fibers will be continuously attenuated while the end of one of the fibers is deflected from the initial position. A micromachined solenoid type inductor with a U-shaped permalloy magnetic core is used to attract the deflectable fiber that has a permalloy coat on its end. To fabricate the multi-layer three-dimensional inductive component, a new UV-LIGA process for thick photoresists is developed, combining advantages of both SU-8 and AZ-4000 series photoresists. The inductive component is approximately 1.7 mm×1.3 mm×50μm in size and has a low resistance value (- 2.1Ω). The whole size of the VOA before packaging is 30 mmx2 mmx0.6 mm. The first prototype shows less then 3-dB insertion loss at 0-dB attenuation and nearly 40-dB attenuation range with less than 20 mW electrical input power at wavelength 1550 nm.展开更多
文摘Giant magneto-impedance(GMI)effect has been realized in the sandwiched FeSiB/Cu/FeSiB films.With magnetic field Ha and ac current applied along the longitudinal direction of the sample,the GMI ratio increases with the increasing Ha,reaching a positive maximum value,and then decreases to negative values with further increase of magnetic field.Field dependence of the GMI ratio also indicates that the magnetic field corresponding to the maximum GMI ratio is different for various frequencies.The positive maximum GMI ratio is 17.2%for Ha=1600A/m and frequency of 3 MHz.In addition,the films display a large negative GMI ratio with a magnetic field applied along the transverse direction and the value of the GMI ratio is about-13.4%for Ha=5600A/m and frequency of 3 MHz.
文摘Coupled Hidden Markov Model (CHMM) is the extension of traditional HMM, which is mainly used for complex interactive process modeling such as two-hand gestures. However, the problems of finding optimal model parameter are still of great interest to the researches in this area. This paper proposes a hybrid genetic algorithm (HGA) for the CHMM training. Chaos is used to initialize GA and used as mutation operator. Experiments on Chinese TaiChi gestures show that standard GA (SGA) based CHMM training is superior to Maximum Likelihood (ML) HMM training. HGA approach has the highest recognition rate of 98.0769%, then 96.1538% for SGA. The last one is ML method, only with a recognition rate of 69.2308%.
基金This project is supported by the National Natural Science Foundation of China (No. 69889050)
文摘A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The output of SVM is moderated to be probability output, which replaces the Mixture of Gauss (MOG) in HMM. Wavelet transformation is used to extract observation vector, which reduces the data dimension and improves the robustness.The hybrid system is compared with pure HMM face recognition method based on ORL face database and Yale face database. Experiments results show that the hybrid method has better performance.
文摘Both sputtering conditions and crystallizing temperatures have great influence on the microstructures and phase transformation characteristics for Ti 51 Ni 44Cu 5. By means of the resistance temperature measurement, X ray diffraction and atomic fore microscopic study, the results indicate that the transformation temperatures of the thin films increase and the "rock candy" martensitic relief is more easily obtained with promoting the sputtering Ar pressure, sputtering power, or crystallizing temperature. However, when sputtering Ar pressure, sputtering power, or crystallizing temperature are lower, a kind of "chrysanthemum" relief, which is related with Ti rich GP zones, is much easier to be observed. The reason is that during crystallization process, both of the inherent compressive stresses introduced under the condition of higher sputtering pressure or higher crystallizing temperature are helpful to the transition from GP zones to Ti 2(NiCu) precipitates and the increase of the transformation temperatures. The addition of copper to substitute for 5% nickel in mole fraction can reduce the transformation hysteresis width to about 10 ～ 15?℃.
文摘The application of reinforcement learning is widely used by multi-agent systems in recent years. An agent uses a multi-agent system to cooperate with other agents to accomplish the given task, and one agent′s behavior usually affects the others′ behaviors. In traditional reinforcement learning, one agent takes the others location, so it is difficult to consider the others′ behavior, which decreases the learning efficiency. This paper proposes multi-agent reinforcement learning with cooperation based on eligibility traces, i.e. one agent estimates the other agent′s behavior with the other agent′s eligibility traces. The results of this simulation prove the validity of the proposed learning method.
文摘This paper presents temporal memory reinforcement learning for the autonomous micro-mobile robot based-behavior. Human being has a memory oblivion process, i.e. the earlier to memorize, the earlier to forget, only the repeated thing can be remembered firmly. Enlightening forms this, and the robot need not memorize all the past states, at the same time economizes the EMS memory space, which is not enough in the MPU of our AMRobot. The proposed algorithm is an extension of the Q-learning, which is an incremental reinforcement learning method. The results of simulation have shown that the algorithm is valid.
文摘A novel MEMS variable optical attenuator (VOA), which has completely different attenuation mechanism from those in literatures, is proposed and demonstrated in this paper. The basic operation principle is that the optical power coupled between two initially aligned single-mode fibers will be continuously attenuated while the end of one of the fibers is deflected from the initial position. A micromachined solenoid type inductor with a U-shaped permalloy magnetic core is used to attract the deflectable fiber that has a permalloy coat on its end. To fabricate the multi-layer three-dimensional inductive component, a new UV-LIGA process for thick photoresists is developed, combining advantages of both SU-8 and AZ-4000 series photoresists. The inductive component is approximately 1.7 mm×1.3 mm×50μm in size and has a low resistance value (- 2.1Ω). The whole size of the VOA before packaging is 30 mmx2 mmx0.6 mm. The first prototype shows less then 3-dB insertion loss at 0-dB attenuation and nearly 40-dB attenuation range with less than 20 mW electrical input power at wavelength 1550 nm.
