The paper focuses on a direct adaptive control plantdeveloped for highly uncertain nonlinear systems, that does notrely on state estimation. In particular, we considersingle-input/single-output nonlinear system, whose...
详细信息
The paper focuses on a direct adaptive control plantdeveloped for highly uncertain nonlinear systems, that does notrely on state estimation. In particular, we considersingle-input/single-output nonlinear system, whose output hasknown, but otherwise arbitrary relative degree. This includessystems with both parameter uncertainty and unmodeleddynamics. The result is achieved by extending the universalfunction approximation property of parameterized neuralnetworks to model unknown system dynamics from input/outputdata. The network weight adaptation rule is derived fromLyapunov function analysis, and guarantees that the adaptedweight errors and the tracking error are bounded. Based on thedesign of adaptive neural network control, a practicalapplication to the Vibroseis system has been achieved.
In this paper, existing iterative multiuser detection tech-niques, including the iterative (Turbo) soft interferencecancellation, the iterative improved parallel interferencecancellation, the gradient descent algorith...
详细信息
In this paper, existing iterative multiuser detection tech-niques, including the iterative (Turbo) soft interferencecancellation, the iterative improved parallel interferencecancellation, the gradient descent algorithm and theslowest-descent search algorithm, are studied. Two newiterative multiuser detection methods, namely, the simpli-fied iterative soft interference cancellation, and the iter-ative slowest-ascent search algorithm, are developed andare compared with the above techniques. It is seen thatthe simplified iterative multiuser detection method signifi-cantly outperforms the other methods in coded CDMA sys-tems except the Turbo receiver, in additive white Gaussianchannels, and also it has low complexity.
For the hardness of diagnosing a complexsystem, the diagnosis often includes many cycles in order to getthe enough diagnostic depth, and the sequential diagnosticstrategy is used widely to enhance the efficiency of th...
详细信息
For the hardness of diagnosing a complexsystem, the diagnosis often includes many cycles in order to getthe enough diagnostic depth, and the sequential diagnosticstrategy is used widely to enhance the efficiency of this *** this paper, an on-line diagnostic algorithm is used in ourexpert system to make the sequential diagnostic decision, andthe multi-signal dependency model is apphed to get therelationship of faults and tests The MADM method based ontile value function is introduced as the evaluation function inthis algorithm for more generalization. Each test is evaluatedsynthetically with two factors, the information gain and cost.A method is proposed to calculate the attribute weights inMADM according to the practical constrained conditions.
The transient performance improvement for anindirect field-oriented (IFO) induction motor drive withfield-weakening is studied in this paper In the proposedmethod, the saturation effect of magnetizing inductance iscon...
详细信息
The transient performance improvement for anindirect field-oriented (IFO) induction motor drive withfield-weakening is studied in this paper In the proposedmethod, the saturation effect of magnetizing inductance isconsidered by representing it using a fitted quadraticpolynomial of flux current. The flux current command iscomposed of three parts, namely the no-load, the loadingcompensated and the transient compensated *** first one is represented by a third-order polynomial ofrotor speed, which is estimated at no-load to consider theeffects of nonlinear magnetizing inductance and thelimitation of motor rated voltage. As for the latter twocomponents, they are utilized to further weaken the fieldsuch that current tracking performances during loadedcases and transient period can be improved.
The design algorithm of computer hardwarefunctional units constitutes a major step in the overalldesign of the system. It requires highly reliablemodeling entities to illustrate the different intermediatestages to go ...
详细信息
The design algorithm of computer hardwarefunctional units constitutes a major step in the overalldesign of the system. It requires highly reliablemodeling entities to illustrate the different intermediatestages to go from problem specification to the finalsystem development. These stages constitute thebehavioral as well as the functional aspect of theoverall system. We illustrate the use of the unifiedmodeling language, UML, using a simplecominunication service and protocol. The differentUML diagrams will be used to configure the systemmodel. In particular, we start with the use casediagran and proceed to the final stage of thedeployment stage A parallel to serial protocolconverter is used to illustrate the method.
This paper proposes using recursive rootmoment method (RRMM) based on feedforward neuralnetworks (FNN) trained by constrained learningalgorithm (CLA) to find the roots of polynomials,which is of fewer computational co...
详细信息
This paper proposes using recursive rootmoment method (RRMM) based on feedforward neuralnetworks (FNN) trained by constrained learningalgorithm (CLA) to find the roots of polynomials,which is of fewer computational complexity than theroot moment method (RMM) and the method using therelations between the roots and the coefficients (RRC)of polynomials. As a result, the RRMM is of fastertraining speed and higher accuracy than the latter twomethods. The experimental results verify our claims.
Speed-stabilization control of a 30,000 rpm high-speed PMSM (permanent magnat synchronous motor) isachjeved without speed sensors. The rotor position of the PMSMis detected by a discrete Hall sensor with the resolutio...
详细信息
Speed-stabilization control of a 30,000 rpm high-speed PMSM (permanent magnat synchronous motor) isachjeved without speed sensors. The rotor position of the PMSMis detected by a discrete Hall sensor with the resolution of ±90electrical degrees. Circuits based on Phase Locked Loop (PLL)are designed to obtain the speed signal and high-precision rotorposition signal. Two filters with different constant time are de-signed to improve the dynamic performance of the-speed-stabilization system because the low-pass filter is adapted in thespeed loop. Experimental and simulated results show the validityof the proposed method.
In this paper,a novel personalized feature combinationscheme is proposed for face recognition. ANFIS (AdaptiveNeuro-Fuzzy Inference System) is adopted to formspecialized feature representation for each subject byfusin...
详细信息
In this paper,a novel personalized feature combinationscheme is proposed for face recognition. ANFIS (AdaptiveNeuro-Fuzzy Inference System) is adopted to formspecialized feature representation for each subject byfusing global and local features. For global features, wemake a comparison between the two traditional globalfeature extraction schemes: PCA and LDA. The localfeatures are extracted with wavelet packet decompositionaround the areas of facial features. Instead of the commonway for different subjects, we realize a new representationthat adapts to each individual. Such adaptability in featureselection is inspired by the face recognition mechanism ofthe human visual system and results in an improvedrecognition rate.
In this study, results obtained by Veil-lette (1995) for continuous systems are extended tothe study of Linear-Exponential-Quadratic-Gaussian(LEQG) type reliable control for discrete-time systemswhich can tolerate abn...
详细信息
In this study, results obtained by Veil-lette (1995) for continuous systems are extended tothe study of Linear-Exponential-Quadratic-Gaussian(LEQG) type reliable control for discrete-time systemswhich can tolerate abnormal operation within somepre-specified set of actuators. This is achieved bysuitable modification of the Algebraic Riccati Equa-tion (ARE) for the design of the controller. Using theLEQG controller's Kalman gain, we show that the sta-bility of the overall system is preserved despite the ab-normal operation of actuators within a pre-specifiedsubset. The gain margin of the feedback gain for re-liable stabilization is also derived, which is shown toagree with the one obtained by Veillette in 1995 forthe continuous case when the sampling time is suffi-ciently small.
A novel power supply system for railwaytraction is present, in which an active power filters isutilized to counteract the quality deteriorating in the threephase power distribution system caused by thesingle-phase tra...
详细信息
A novel power supply system for railwaytraction is present, in which an active power filters isutilized to counteract the quality deteriorating in the threephase power distribution system caused by thesingle-phase traction system. The proposed system adoptsan uniphase topology, which enables all the trains run onthe same phase along its whole journey whereas, at everypoints where the two system are connected, thesingle-phase load acts as a symmetrical pure resistant ofthe three-phase system. Configuration and analysis of thesystem, a frequency-limited hysteresis current controlstrategy, and software simulation results that can prove thesystem's feasibility are included.
暂无评论