Advanced Materials Research Vols. 875-877

Abstract: For robot application to proliferate in industry, and in unregulated environments, a simple means of programming is required. This paper describes methods for robot Learning from Demonstration (LfD). These methods used an RGB-D sensor for demonstration observation, and used finite state machines (FSMs) for policy derivation. Particularly, a method for object recognition was developed, which required only a single frame of data for training, and was able to perform real-time recognition. A planning method for object grasping was also developed. Experiments with a pick-and-place robot show that the developed methods resulted in object recognition accuracy greater than 99% in cluttered scenes, and manipulation accuracies of below 3mm in linear motion and 2° in rotation.

Abstract: A procedure to calculate the transient response of optimal controlled stochastic Van Der Pol oscillator is proposed. The stochastic averaging method is employed to obtain a partially averaged Itô equation for the amplitude process. The dynamical programming equation is adopted to minimize the system response. An optimal control law with a control constraint is established. The completed averaged Itô equation is obtained. The transient probability density function is solved from Fokker-Planck-Kolmogorov equation by Galerkin method. Results obtained show the proposed method is accurate. The effective of the control strategy is significant.

Abstract: In this paper, a novel technology, which detects and forecasts human's abnormal behavior, is proposed. Taking an abnormal behavior, the robbery, as an example, the accuracy of the technology is demonstrated successfully. There are 3 innovations in this paper. A) A robbery incident is divided into 5 phases occurred orderly. By its order, a robbery is detected. B) By human body's motion character, the human body's speeds are measured in different areas and at different times. It improves the accuracy rate of a classifier. C) The technology can forecast an abnormal human behavior in a short time advance. Experiment shows that the technology detecting and forecasting accuracy are high enough to be used in practice.

Abstract: In this paper a higher order sliding mode control based on the twisting algorithm is studied. The aim is to prove the impact of the choice of sliding surface in the design of the control law. The simulation results shows, a proper selection of the sliding surface in the design of the control law can eliminate or reduce the phenomenon of reluctance. The phenomenon of reluctance or chattering is a major drawback significant in sliding mode control because, although it is possible to filter the output of the process, it is likely to excite high frequency modes that have not been included in the model of the system. This can degrade performance and even lead to instability. For the design of the control law of vehicle model, two sliding surfaces are integrated; the first show a good tracking in the simulation result with disadvantage of the chattering presence in the control law while the second provide a good performance without chattering.

Abstract: High dimensional data such as mass-spectrometric and near-infrared spectrum are always used in disease diagnosis and product quality monitoring. Aim at the nonlinear feature extraction and low learning speed problems, a novel modeling approach combined principal component analysis (PCA) with kernel extreme learning machine (KELM) is proposed. The extracted features using PCA algorithms are fed into nonlinear classification based KELM with fast learning speed. The numbers of the features are selected according the classification performance. The experimental results based on the mass-spectrometric data in the benchmark demonstrate that the proposed approach has better performance. This approach can also be used to target recognition based on radar data.

Abstract: Feature selection technology can improve the modeling accuracy and reduce model’s complexity, especially for the high dimensional spectral data. Aim at this problem, feature selection approach based on mutual information (MI) and partial least square (PLS) is proposed in this paper. MI values between features and responsible variable are calculated, and the threshold value using to select final features is optimal selected based on PLS algorithm. The numbers of the latent values of the PLS and the threshold value of MI are selected according the modeling performance simultaneously. The experimental results based on the near-infrared spectrum show that the proposed approach has better performance.

Abstract: Due to the parametric and structural uncertainty of the DC drive system, an adaptive control method is necessary. Therefore, an original model reference adaptive control (MRAC) for DC drives is proposed in this paper. MRAC ensures on-line adjustment of the control parameters with DC machine parameter variation. The proposed adaptive control structure provides regulating advantages: asymptotic cancellation of the tracking error, fast and smooth evolution towards the origin of the phase plan due to a sliding mode switching k-sigmoid function. The reference model can be a real strictly positive function (the tracking error is also the identification error) as its order is relatively higher than one degree. For this reason, the synthesis of the adaptive control will use a different type of error called augmented or enhanced error. The DC machine with separate excitation is fed at a constant flux. This adaptive control law assures robustness to external perturbations and to unmodelled dynamics.

Abstract: This paper discussed the possibility of deploying the proposed Variable Geometry Tracked Vehicle (VGTV) in our first robotic based prototype called REC-01. REC-01 is being developed as a miniature scouting and search & rescue all-terrain robotic vehicle. Its comparatively small size and mass limit its power capacity and navigational capabilities for larger obstacles. The VGTV is a unique concept used to overcome these shortcomings. The main part of this paper is to discuss the mechanical efficiency of the tracked systems in some specific field conditions and compare the results with the conventional tracked propulsion. The outcomes of the first prototype are also briefly discussed in this paper.

Abstract: Reliability growth assessment is an important part of the reliability growth management. According to the characteristic that the lifetime of electromechanical product obeys the Weibull distribution, the thesis provides a monotonous constraint model of reliability growth of Weibull distribution, including complete Samples and censored samples. The method proposed in this paper has been verified by actual data, and has good adaptability. The comprehensive evaluation results of MTBF are higher than those got through only utilizing the test data of the last stage.Background Information

Abstract: according to the requirement of designing the magnetic sealing drive, the magnetic sealing drive torque design and calculation is completed. It is also analyzed the magnetic circuit and the whole structure of the drive. The matching problem of magnet thickness and arc length of magnetic pole surface is analyzed. The research also includes the method to calculate torque of combination of push pull magnetic circuit and method of its power matching calculation and correction. The magnetic sealing drive is tested and the results show that the design of the magnetic sealing drive complies with the design requirements.