Variable Perturbation Size Adaptive P&O MPPT Algorithm for Sudden Changes in Irradiance

2014 ◽  
Vol 5 (3) ◽  
pp. 718-728 ◽  
Author(s):  
Sathish Kumar Kollimalla ◽  
Mahesh Kumar Mishra
Keyword(s):  
Author(s):  
Weiya Jin ◽  
Brian H. Dennis ◽  
Bo Ping Wang

In fracture mechanics, due to singularity existing on the crack tip, the corresponding sensitivity analysis of Stress Intensity Factor (SIF) is not clear whether the overall finite difference (OFD) or Semi-analytical Method (SAM) can obtain accurate sensitivities. The paper proposes the Semi-analytical Complex Variable Method (SACVM) to compute sensitivities of SIF and compares the sensitivities computed by the SACVM with those computed by the OFD and SAM in Center Cracked Tension (CCT) specimen and Single Edge Notched Tension (SENT) specimen. The results reveal that the OFD obtains oscillated sensitivities because of the ill-conditioned linear system. The sensitivities computed by the OFD and SAM are sensitive to the perturbation size out of a certain range. However, this certain range varies with different variable, and is not known a priori. The proposed SACVM can always obtain accurate, consistent sensitivities with little extra computational cost than the SAM. The SACVM is not sensitive to the perturbation size and is not affected by the ill-conditioned linear system. Therefore, the SACVM is recommended to deal with sensitivity analysis in the fracture mechanics.


Author(s):  
C. Alippi

This chapter presents a general methodology for evaluating the loss in performance of a generic neural network once its weights are affected by perturbations. Since weights represent the “knowledge space” of the neural model, the robustness analysis can be used to study the weights/performance relationship. The perturbation analysis, which is closely related to sensitivity issues, relaxes all assumptions made in the related literature, such as the small perturbation hypothesis, specific requirements on the distribution of perturbations and neural variables, the number of hidden units and a given neural structure. The methodology, based on Randomized Algorithms, allows reformulating the computationally intractable problem of robustness/sensitivity analysis in a probabilistic framework characterised by a polynomial time solution in the accuracy and confidence degrees.


2015 ◽  
Vol 28 ◽  
pp. 114-122 ◽  
Author(s):  
Laurence Yang ◽  
Shyamsundhar Srinivasan ◽  
Radhakrishnan Mahadevan ◽  
William R. Cluett

Author(s):  
Sogkyun Kim ◽  
Sean Ellis ◽  
Mark Challener

Real-Time Engine Models are required for operation with engine electronic control systems and/or aircraft simulators for functional demonstration. The challenge for Rolls-Royce has been to establish the sub-idle speed behaviour of the engine. This paper covers the development steps by the Civil Aerospace Modelling and Simulation team to resolve this limitation in the models. The real-time engine model is now generated using two non-linear thermodynamic engine models. One of the thermodynamic engine models, normal range, covers the idle to max power range and the other is for sub-idle operation. Previously sub-idle operation was established by extrapolation from the normal range model. However, this method limited control system development by simulation for altitude starting adding time to altitude test programmes in high cost facilities. The requirement for the technique is to obtain the partial derivatives and steady-state data for the whole operating range. For the partial derivative estimation in sub-idle region, a variable perturbation size is introduced and changed according to the different shaft speed so that the sensitivity issue of using a fixed perturbation size in this operating range is resolved. Furthermore, the partial derivative of each parameter from the non-linear models is fine tuned by comparing with the steady-state values for each parameter. The summation of the integrated partial derivatives should be same as the steady-state value of each engine parameter. If an error exists then an adjustment of each integrated partial derivative is conducted according to the relative weight of each integrated partial derivatives contribution to the whole. It is highlighted that error sharing between the integrated partial derivative parameters results in less error during the validation process. The real-time engine model is constructed in state-space modular subsystems in SIMULINK, which include an engine shaft block to generate the engine shaft speeds, and fuel block to generate a signal of engine lit, etc. The database generated by the process of partial derivatives is then used in calculation of engine’s shaft speeds, temperatures and pressures. For the test of the real-time engine model obtained in this study, simulation of engine starting from stationary is conducted. Using a starter torque as the input to the engine model, starter-assisted starting can be achieved. In addition, engine relighting in flight is also conducted. The output of the real-time engine model has been compared with flight test data for engine relight and agreement has been demonstrated.


2014 ◽  
Vol 748 ◽  
pp. 521-548 ◽  
Author(s):  
Clara O’Farrell ◽  
John O. Dabiri

AbstractInviscid models for vortex rings and dipoles are constructed using nested patches of vorticity. These models constitute more realistic approximations to experimental vortex rings and dipoles than the single-contour models of Norbury and Pierrehumbert, and nested contour dynamics algorithms allow their simulation with low computational cost. In two dimensions, nested-contour models for the analytical Lamb dipole are constructed. In the axisymmetric case, a family of models for vortex rings generated by a piston–cylinder apparatus at different stroke ratios is constructed from experimental data. The perturbation response of this family is considered by the introduction of a small region of vorticity at the rear of the vortex, which mimics the addition of circulation to a growing vortex ring by a feeding shear layer. Model vortex rings are found to either accept the additional circulation or shed vorticity into a tail, depending on the perturbation size. A change in the behaviour of the model vortex rings is identified at a stroke ratio of three, when it is found that the maximum relative perturbation size vortex rings can accept becomes approximately constant. We hypothesise that this change in response is related to pinch-off, and that pinch-off might be understood and predicted based on the perturbation responses of model vortex rings. In particular, we suggest that a perturbation response-based framework can be useful in understanding vortex formation in biological flows.


Sign in / Sign up

Export Citation Format

Share Document