Catálogo de publicaciones - libros

In this paper a fuzzy logic controller to minimize the torque ripples associated wit DTC (FL_DTC) of PMSM is presented. In this system, in order to attain the merits of fast torque response, the torque error, the flux error and stator flux position sector are used to select one active vector while the active switching time of this vector is selected as a result of the torque error magnitude and flux error magnitude fuzzification. The output of such fuzzy logic controller is used to adapt the inverter switching to follow the errors magnitude level. Arranging the inverter switching in this manner, results in fast torque response and minimum torque ripples as compared to the traditional HDTC. Additionally it results in less switching losses associated with SVMDTC. The simulated results show considerable torque ripple as well as current ripple reduction. The simulated results are supported with experimental results.

Based on the modified ant colony system (ACS) algorithm, a novel design method for nonlinear PID controller with on-line optimal self-tuning gains is proposed. In this method, first the regulative laws of the three PID controller gains are designed respectively, which are three nonlinear functions of system error and its variation rate, and then the proposed modified ACS algorithm is used to optimize the nine parameters in the three nonlinear functions. The designed controller is called the ACS-NPID controller and was used to control the CIP-I intelligent leg prosthesis. The simulation experiments demonstrated that the ACS-NPID controller has better control performance compared with other three linear PID controllers designed respectively by the differential evolution algorithm, the real-coded genetic algorithm, and the simulated annealing. The simulation results also verified that the modified ACS algorithm has good performance in convergence speed and solution variation characteristic.

A Neural Network based AC-AC voltage restorer is proposed for voltage sags and PWM type active power filter is used for voltage harmonics compensation and EMI reduction. The objective is to apply the neural network switching control technique to the AC-AC voltage restorer to reduce time delays during the switching conditions and switching losses. The aim of the IGBTs used in the AC-AC voltage restorer is to test and to find the best switching frequency-power combination in the steps of the simulation. Thus, the proposed AC-AC voltage restorer has some advantages such as fast switching response, simplicity and more intelligent structure, better output waveform. Neural Network techniques have proven that they were suitable for parameter identification and control of nonlinear systems. The transient condition of the AC-AC voltage restorer is improved via the Neural Network based control technique. On the other side, the proposed strategy for elimination of voltage harmonics using PWM type DC-AC inverter part of the system as an active filter. The last objective of the system is ElectroMagnetic Interference (EMI) reduction with using this filter and voltage restorer.

This paper proposes obstacle collision avoidance algorithm for a mobile robot. To navigate without colliding with obstacles the mobile robot uses both the vision system approach and edge detection approach using ultrasonic sensor. The vision system approach is used when the obstacle is located more than 2 meters away from the mobile robot, and the edge detection approach using ultrasonic sensor is used for detecting obstacle remains within 2 meters range. The collision avoidance algorithm proposed in this paper utilizes both the limit-cycle method and the nearness diagram navigation method. The proposed method is tested using Pioneer 2-DX mobile robot. Through experiment, the effectiveness of proposed methods for navigating a mobile robot in the dynamic environment is demonstrated.

In this paper, identification and adaptive control of networked control systems (NCSs) with uncertain networked-induced delay is considered. A novel and useful scheme for simultaneous estimation of networked-induced delays and dynamics of the plant is presented using linear filter method. Based on the suggested identification algorithm, a quadratically stabilizable adaptive control law with guaranteed cost is derived.

To solve a problem with intelligent planning, an expert has to try his best to write a planning domain. It is hard and time-wasting. Considering software requirement as a problem to be solved by intelligent planning, it’s even more difficult to write the domain, because of software requirement’s feature, for instance, changeability. To reduce the difficulty, we divide the work into two tasks: one is to describe an incomplete domain of software requirement with PDDL(Level 1,Strips) [11]; the other is to complete the domain by learning from plan samples based on business processes. We design a learning tool (Learning Action Model from Plan Samples, LAMPS) to complete the second task. In this way, what an expert needs to do is to do the first task and give some plan samples. In the end, we offer some experiment result analysis and conclusion.

The motion identification for a class of movements in the space by using stereo vision is considered by observing at least three points in this paper. The considered motion equation can cover a wide class of practical movements in the space. The observability of this class of movement is clarified. The estimations of the motion parameters which are all time-varying are developed in the proposed algorithm based on the second method of Lyapunov. The assumptions about the perspective system are reasonable, and the convergence conditions are intuitive and have apparently physical interpretations. The proposed recursive algorithm requires minor a priori knowledge about the system and can alleviate the noises in the image data. Furthermore, the proposed algorithm is modified to deal with the occlusion phenomenon. Simulation results show the proposed algorithm is effective even in the presence of measurement noises.

This paper presents a systematic isotropy analysis of a caster wheeled omnidirectional mobile robot (COMR) with three active caster wheels. Unlike previous analysis, no assumption is made on the relative scale of the steering link offset and the wheel radius. First, with the characteristic length introduced, the kinematic model of a COMR is obtained based on the orthogonal decomposition of the wheel velocities. Second, the necessary and sufficient isotropy conditions are examined to categorize three different groups to be handled in a similar way. Third, the isotropy conditions are further explored to identify four different sets of all possible isotropic configurations. Fourth, the characteristic lengths required for the isotropy of a COMR are obtained in a closed-form. Finally, the local and the global isotropy indices are used to determine the optimal design parameters.

This paper describes a control strategy for mobile robot Auto-seeking the Welded joint. Through this control strategy the mobile robot can find position of the weld joint automatically and adjust welding torch to parallel to it. Many simulation experiments results demonstrate the effectiveness and correctness of the control algorithm.

In the transient stability analysis of power systems it is extremely important to determine the boundary of stability region. Owing to the complexity and multi-time scale natures of electric power systems, it is necessary to correct boundary of stability region of simplified system (approximate reduction order system). In this paper, determining conditions of the stability boundary of multi-time scale systems is presented, O(ε)-correction formula of the stability boundary is obtained for multi-time scale power systems, which decreases the errors caused by employing the approximate reduced model to instead of the exact reduced system. An example and a simulation of one-machine infinite-bus electric power system are given to illustrate the validity of the O(ε)-correction algorithm.