Catálogo de publicaciones - libros

In this paper a new method is presented to solve a series of fault detection problems using 3 σ rule in Genetic Programming (GP). Fault detection can be seen as a problem of multi-class classification. GP methods used to solve problems have a great advantage in their power to represent solutions to complex problems and this advantage remains true in the domain of fault detection. Moreover, diagnosis accuracy can be improved by using 3 σ rule. In the end of this paper, we use this method to solve the fault detection of electro-mechanical device. The results show that the method uses GP with 3 σ rule to solve fault detection of electro-mechanical device outperforms the basic GP and ANN method.

Nowadays, empirical mode decomposition (EMD) and Hilbert spectrum (HS) have been broadly applied on non-stationary and nonlinear signal analysis because it is developed from instantaneous frequency. HS offers a means of signal pattern recognition. But it is the same as other time-frequency distribution methods without giving a quantitative description. Thus, it is not very suitable for real application. In this research, pattern recognition method based on Kolmogorov-Smirnov (K-S) test method is investigated in detail for machine condition monitoring and pattern recognition. For accurate and automatic recognition for the working of machines, K-S test is used to distinguish and recognize the engine condition. Experimental data of a marine diesel engine is used to evaluate the developed methodology for diesel automatic pattern recognition. It can be concluded that this method is very promising for the development of diesel engine preventative maintenance.

The empirical mode decomposition (EMD) is a new method for adaptive decomposition of nonlinear and non-stationary signals. This paper combines the EMD method and the conventional response modulation method to analyze the fault in the rolling bearing. B-spline EMD is also investigated to extract the high frequency components of the signal. Then we use discrete time Fourier transform (DTFT) to get the refinement spectrum analysis. The results show that the combination of these two methods performs excellently.

A QoS resource error detection-recovery model called “AFQ_WE” was proposed for situation-aware middleware as RCSM. AFQ_WE model was used to detect and recover the QoS resource errors among actions. An example of situation-aware applications is a multimedia education system. Education system for distributed multimedia holds the promise of greatly improving all forms of remote education and training. The model aims at guaranteeing it through application QoS. The AFQ_WE model is proposed for supporting QoS resource errors detection-recovery in the situation-aware middleware.

This paper proposed a non-linear signal processing method to separate the stress wave signals from different sources for extracting the multi-fault features without any a priori knowledge. Firstly, stress wave theories of low-speed rotating machines are discussed. And then this paper described the basic principles of BSS, and a kind of non-linear neural network is used in consideration of the complex behavior of vibration, either nonlinear or random. After that BSS was applied to the experimental simulations for diagnosing the rubbing and looseness defects. Finally the processing results of a real process show that the new method based on non-linear blind source separation studied in this paper can simply and effectively remove the noise contained in the vibration signals and extract the multi-fault features to avoid the grand accidence.

The artificial potential field (APF) based path planning methods have a local minimum problem, which can trap mobile robot before reaching its goal. In this study, a new method using virtual water-flow is proposed to escape local minima occurred in local path planning, which integrates virtual water-flow with a potential-field-based method to guide a mobile robot in an unknown or unstructured environment. The potential-field method coupled with virtual water-flow can navigate a mobile robot in real time. Simulations and experiments show this algorithm possesses good performance, and can overcome the problem cause by local minimum.

This paper presents a novel and efficient method for autopilot design by integrating the Particle Swarm Optimization (PSO) algorithm with the Sequential Quadratic Programming (SQP) technique. In the proposed hybrid PSO-SQP algorithm, PSO algorithm is the basic optimizer and the SQP technique is used to reduce computation time and to improve convergence performance. Also, the proposed hybrid PSO-SQP algorithm is applied to autopilot design for a transport aircraft and an air-to-air missile. In view of autopilot design, we note that the proposed design method has further flexibility and control performance than classical method.

Aproposed sliding mode controller gain has been presented. The traditional switching function smooth method is improved by combining the smoothed switching function with the time-varying control gain. And we provide a sliding mode controller. The proposed controller has been applied to a motion trajectory tracking of mobile robot. Simulation study has been carried out to evaluate the performance of the proposed controller and to compare its performance with a recent conventional sliding mode controller performance. The simulation results showed that the proposed sliding mode controller has impressively achieved a highly performance with minimum reaching time and smooth control actions. In addition the results have proven the effectiveness and robustness of the proposed controller to achieve considerable enhancement of the motion tracking of robotics characterized by disturbances and parameter uncertainties without any obvious chattering.

Dependable distributed embedded systems (DDES) play an increasing role in modern automotive designs and there is a pressing need to evaluate the effect of different design system architecture on system safety and reliability. Both hardware and software architecture can impact system dependability. The evaluation of a dependable system is based on vehicle control system, which is to perform testing and assessment for various system architecture. The evaluation embeds vehicle control hardware and operational environments. Different software architectures such as time-triggered, event-triggered and hybrid-triggered were tested during the evaluation process. In the paper an overview of the simulator is provided, followed by detailed descriptions of the vehicle and driver models that are employed. Simulation show these models comply with manufacturer and empirical data. The performance of the evaluation process is illustrated using an example based on an adaptive cruise control system (ACCS).

In this paper, the passivity property is applied to stability analysis, controller synthesis and observer design of the discrete affine Takagi-Sugeno (T-S) fuzzy systems. A fuzzy controller is designed to achieve Lyapunov stability of system and strictly input passivity for all exogenous disturbances. The proposed fuzzy controller design approach is developed based on the Parallel Distribution Compensation (PDC) technique. Moreover, an Iterative Linear Matrix Inequality (ILMI) algorithm is applied to solve the feasible solutions for the proposed observer feedback fuzzy control problem. Finally, the application of present approach is illustrated by a numerical example.