Harr Wavelet Research Articles

Effective Energy Management Strategy with Model-Free DC-Bus Voltage Control for Fuel Cell/Battery/Supercapacitor Hybrid Electric Vehicle System

This article presents a new design method of energy management strategy with model-free DC-Bus voltage control for the fuel-cell/battery/supercapacitor hybrid electric vehicle (FCHEV) system to enhance the power performance, fuel consumption, and fuel cell lifetime by considering regulation of DC-bus voltage. First, an efficient frequency-separating based-energy management strategy (EMS) is designed using Harr wavelet transform (HWT), adaptive low-pass filter, and interval type–2 fuzzy controller (IT2FC) to determine the appropriate power distribution for different power sources. Second, the ultra-local model (ULM) is introduced to re-formulate the FCHEV system by the knowledge of the input and output signals. Then, a novel adaptive model-free integral terminal sliding mode control (AMFITSMC) based on nonlinear disturbance observer (NDO) is proposed to force the actual values of the DC-link bus voltage and the power source’s currents track their obtained reference trajectories, wherein the NDO is used to approximate the unknown dynamics of the ULM. Moreover, the Lyapunov theorem is used to verify the stability of AMFITSMC via a closed-loop system. Finally, the FCHEV system with the presented method is modeled on a Matlab/Simulink environment, and different driving schedules like WLTP, UDDS, and HWFET driving cycles are utilized for investigation. The corresponding simulation results show that the proposed technique provides better results than the other methods, such as operational mode strategy and fuzzy logic control, in terms of the reduction of fuel consumption and fuel cell power fluctuations.

Data-augmented wavelet capsule generative adversarial network for rolling bearing fault diagnosis

Rolling bearing fault diagnosis with limited imbalance data is significant and challenging. It is​ a nice attempt to generate data for balancing datasets. In this paper, a wavelet capsule generative adversarial network (WCGAN) is proposed to address this issue. Firstly, the Harr wavelet is introduced into GAN to construct wavelet transform GAN (WTGAN). It keeps convolutional neural networks (CNNs) shift-invariant to extract the deep features of the data. Secondly, WCGAN is developed to alleviate CNNs’ incomplete analysis of signal information, which replaces part of CNNs in WTGAN with capsule networks. Thirdly, a novel loss function is designed for WCGAN to maintain a smooth training process and improve the quality of the generated data. Furthermore, various experiments are conducted in multiple ways to confirm the effectiveness and accuracy of the novel method. Results indicate that the proposed method balances the dataset and surpasses other advanced approaches in imbalanced data diagnosis with potential.

An Efficient Methodology for Brain MRI Classification Based on DWT and Convolutional Neural Network.

In this paper, a model based on discrete wavelet transform and convolutional neural network for brain MR image classification has been proposed. The proposed model is comprised of three main stages, namely preprocessing, feature extraction, and classification. In the preprocessing, the median filter has been applied to remove salt-and-pepper noise from the brain MRI images. In the discrete wavelet transform, discrete Harr wavelet transform has been used. In the proposed model, 3-level Harr wavelet decomposition has been applied on the images to remove low-level detail and reduce the size of the images. Next, the convolutional neural network has been used for classifying the brain MR images into normal and abnormal. The convolutional neural network is also a prevalent classification method and has been widely used in different areas. In this study, the convolutional neural network has been used for brain MRI classification. The proposed methodology has been applied to the standard dataset, and for performance evaluation, we have used different performance evaluation measures. The results indicate that the proposed method provides good results with 99% accuracy. The proposed method results are then presented for comparison with some state-of-the-art algorithms where simply the proposed method outperforms the counterpart algorithms. The proposed model has been developed to be used for practical applications.

An artificial intelligence-driven agent for real-time head-and-neck IMRT plan generation using conditional generative adversarial network (cGAN).

To develop an artificial intelligence (AI) agent for fully automated rapid head-and-neck intensity-modulated radiation therapy (IMRT) plan generation without time-consuming dose-volume-based inverse planning. This AI agent was trained via implementing a conditional generative adversarial network (cGAN) architecture. The generator, PyraNet, is a novel deep learning network that implements 28 classic ResNet blocks in pyramid-like concatenations. The discriminator is a customized four-layer DenseNet. The AI agent first generates multiple customized two-dimensional projections at nine template beam angles from a patient's three-dimensional computed tomography (CT) volume and structures. These projections are then stacked as four-dimensional inputs of PyraNet, from which nine radiation fluence maps of the corresponding template beam angles are generated simultaneously. Finally, the predicted fluence maps are automatically postprocessed by Gaussian deconvolution operations and imported into a commercial treatment planning system (TPS) for plan integrity check and visualization. The AI agent was built and tested upon 231 oropharyngeal IMRT plans from a TPS plan library. 200/16/15 plans were assigned for training/validation/testing, respectively. Only the primary plans in the sequential boost regime were studied. All plans were normalized to 44Gy prescription (2Gy/fx). A customized Harr wavelet loss was adopted for fluence map comparison during the training of the PyraNet. For test cases, isodose distributions in AI plans and TPS plans were qualitatively evaluated for overall dose distributions. Key dosimetric metrics were compared by Wilcoxon signed-rank tests with a significance level of 0.05. All 15 AI plans were successfully generated. Isodose gradients outside of PTV in AI plans were comparable to those of the TPS plans. After PTV coverage normalization, Dmean of left parotid (DAI =23.1±2.4Gy; DTPS =23.1±2.0Gy), right parotid (DAI =23.8±3.0Gy; DTPS =23.9±2.3Gy), and oral cavity (DAI =24.7±6.0Gy; DTPS =23.9±4.3Gy) in the AI plans and the TPS plans were comparable without statistical significance. AI plans achieved comparable results for maximum dose at 0.01cc of brainstem (DAI =15.0±2.1Gy; DTPS =15.5±2.7Gy) and cord+5mm (DAI =27.5±2.3Gy; DTPS =25.8±1.9Gy) without clinically relevant differences, but body Dmax results (DAI =121.1±3.9Gy; DTPS =109.0±0.9Gy) were higher than the TPS plan results. The AI agent needed ~3s for predicting fluence maps of an IMRT plan. With rapid and fully automated execution, the developed AI agent can generate complex head-and-neck IMRT plans with acceptable dosimetry quality. This approach holds great potential for clinical applications in preplanning decision-making and real-time planning.

Measurement report: characteristics of clear-day convective boundary layer and associated entrainment zone as observed by a ground-based polarization lidar over Wuhan (30.5° N, 114.4° E)

Abstract. Knowledge of the convective boundary layer (CBL) and associated entrainment zone (EZ) is important for understanding land–atmosphere interactions and assessing the living conditions in the biosphere. A tilted 532 nm polarization lidar (30∘ off zenith) has been used for the routine atmospheric measurements with 10 s time and 6.5 m height resolution over Wuhan (30.5∘ N, 114.4∘ E). From lidar-retrieved aerosol backscatter, instantaneous atmospheric boundary layer (ABL) depths are obtained using the logarithm gradient method and Harr wavelet transform method, while hourly mean ABL depths are obtained using the variance method. A new approach utilizing the full width at half maximum of the variance profile of aerosol backscatter ratio fluctuations is proposed to determine the entrainment zone thickness (EZT). Four typical clear-day observational cases in different seasons are presented. The CBL evolution is described and studied in four developing stages (formation, growth, quasi-stationary and decay); the instantaneous CBL depths exhibited different fluctuation magnitudes in the four stages and fluctuations at the growth stage were generally larger. The EZT is investigated for the same statistical time interval of 09:00–19:00 LT. It is found that the winter and late autumn cases had an overall smaller mean (mean) and standard deviation (SD) of EZT data compared to those of the late spring and early autumn cases. This statistical conclusion was also true for each of the four developing stages. In addition, compared to those of the late spring and early autumn cases, the winter and late autumn cases had larger percentages of EZT falling into the subranges of 0–50 m but smaller percentages of EZT falling into the subranges of > 150 m. It seems that both the EZT statistics (mean and SD) and percentage of larger EZT values provide measures of entrainment intensity. Common statistical characteristics also existed. All four cases showed moderate variations of the mean of the EZT from stage to stage. The growth stage always had the largest mean and SD of the EZT and the quasi-stationary stage usually the smallest SD of the EZT. For all four stages, most EZT values fell into the 50–150 m subrange; the overall percentage of the EZT falling into the 50–150 m subrange between 09:00 and 19:00 LT was > 67 % for all four cases. We believe that the lidar-derived characteristics of the clear-day CBL and associated EZ can contribute to improving our understanding of the structures and variations of the CBL as well as providing a quantitatively observational basis for EZ parameterization in numerical models.

Groundwater level prediction in arid areas using wavelet analysis and Gaussian process regression

Utilizing new approaches to accurately predict groundwater level (GWL) in arid regions is of vital importance. In this study, support vector regression (SVR), Gaussian process regression (GPR), and their combination with wavelet transformation (named wavelet-support vector regression (W-SVR) and wavelet-Gaussian process regression (W-GPR)) are used to forecast groundwater level in Semnan plain (arid area) for the next month. Three different wavelet transformations, namely Haar, db4, and Symlet, are tested. Four statistical metrics, namely root mean square error (RMSE), mean absolute percentage error (MAPE), coefficient of determination (R 2), and Nah-Sutcliffe efficiency (NS), are used to evaluate performance of different methods. The results reveal that SVR with RMSE of 0.04790 (m), MAPE of 0.00199%, R 2 of 0.99995, and NS of 0.99988 significantly outperforms GPR with RMSE of 0.55439 (m), MAPE of 0.04363%, R2 of 0.99264, and NS of 0.98413. Besides, the hybrid W-GPR-1 model (i.e. GPR with Harr wavelet) remarkably improves the accuracy of GWL prediction compared to GPR. Finally, the hybrid W-SVR-3 model (i.e. SVR with Symlet) provides the best GWL prediction with RMSE, MAPE, R2, and NS of 0.01290 (m), 0.00079%, 0.99999, and 0.99999, respectively. Overall, the findings indicate that hybrid models can accurately predict GWL in arid regions.

A fault analysis for Micro-grid based on Wavelet

Micro grid is an efficient and flexible way to solve the problem of grid-connected distributed generation. Since the distributed generation accessed and the number of branch increases, system fault condition is complicated. Micro grid fault protection becomes a hot issue now. In this paper, a method based on wavelet is proposed for micro grid fault current analysis. A typical 10kV micro grid model is established in PSCAD, which is consisted of wind power, PV, EV and so on. And, short circuit fault is simulated. Then, the fault current data is further dealt by Harr wavelet. The results would provide a good guide for the setting of fault protection scheme.

Frequency Decoupling-Based Energy Management Strategy for Fuel Cell/Battery/Ultracapacitor Hybrid Vehicle Using Fuzzy Control Method

In order to extend fuel cell lifespan and improve fuel economy of electrical hybrid vehicle with fuel cell/battery/ultracapacitor (FCHEV), a frequency decoupling-based energy management strategy (EMS) for FCHEV using fuzzy control method is proposed. In detail, firstly, according to different characteristics of energy sources, required power of FCHEV is decomposed into three frequency ranges based on Harr wavelet transform and an adaptive-fuzzy filter. Secondly, based on the proposed frequency decoupling, the obtained three frequency required power is supplied by fuel cell-battery and ultracapacitor, respectively, which can guarantee power performance of vehicle and reduce pressure and power fluctuation on fuel cell and battery. Thirdly, for improving fuel economy, one fuzzy controller is proposed to split the power between fuel cell and battery. Finally, the proposed strategy in this paper is verified by advisor-simulink and experimental bench. Simulation and experimental results show that the proposed EMS can effectively reduce impact of power fluctuations on fuel cell, extend its lifespan and reduce fuel consumption on 7.94% compared to equivalent consumption minimization strategy.

The Method of Vehicle Type Identification Based on Improved Corner Feature

This paper present a method of vehicle type identification using improved corner points as key points for feature description. The corner points can reflect the information of different type. Therefore, we adopt the features of corner points to recognize which type the vehicle belongs to. The method of Harris corner detection was improved, and the ratio of eigenvalues of M matrix in different directions was added to the corner response. The corner points were detected by improved corner response and results of test showed that the count of the detected corner points was increased. Considering rotation invariance and scale invariance, feature description is carried out around detected corner points. The vector of feature is determined by the Harr wavelet response of the pixels around the corner point, but the direction is only one of the eight values, and the length of the vector is 16. The matching rate of feature vectors between different types is used to determine which type the vehicle belongs to. The results of experiment confirmed the validity of the method.

Efficient oppositional based optimal Harr wavelet for compound image compression using MHE

Image compression is to reduce the size in bytes of a graphics file without corrupting the nature of the image to an unsatisfactory level. The compound image compression ordinarily in view of three classification strategies that is object-based, layer based along with block based. The corruption of compound image or report compression depends on the capacity and transmission of the files. This proposed model has two methods utilized for compression procedure that is optimal harr wavelets and Merging Huffman Encoding (MHE) process. Discrete Wavelet Transform (DWT)-Harr wavelet coefficient optimization analysis encouraged Oppositional based Grey Wolf Optimization (OGWO). Then the compression strategy module based encryption is done. From the outcomes accomplished the highest Compression Ratio (CR) for a compound image that is 97.56% compared with existing systems and furthermore assessed Peak signal to noise ratio (PSNR), Mean square Error (MSE), and Compression Size (CS) of all testing images.

A novel approach framework based on statistics for reconstruction and heartrate estimation from PPG with heavy motion artifacts

One of the most important applications of photoplethysmography (PPG) signal is heartrate (HR) estimation. For its applications in wearable devices, motion artifact (MA) may be the most serious challenge for randomness both in format and temporal distribution. This paper proposes an advanced time-frequency analysis framework based on empirical mode decomposition (EMD) to select specific time slices for signal reconstruction. This framework operates with a type of pre-processing called variance characterization series (VCS), EMD, singular value decomposition (SVD), and a precise and adaptive 2-D filtration reported first. This filtration is based on Harr wavelet transform (HWT) and 3rd order cumulant analysis, to make it have resolution in both the time domain and different components. The simulation results show that the proposed method gains 1.07 in absolute average error (AAE) and 1.87 in standard deviation (SD); AAEs 1st and 3rd quartiles are 0.12 and 1.41, respectively. This framework is tested by the PhysioBank MIMIC II waveform database.

An Efficient Compound Image Compression Using Optimal Discrete Wavelet Transform and Run Length Encoding Techniques

Abstract Reduction in file size leads to reduction in the number of bits required to store it. When data is compressed, it must be decompressed into its original form bit for bit. Compound images are defined as images that contain a combination of text, natural (photo) images and graphic images. Here, compression is the process of reducing the amount of data required to represent information. Image compression is done on the basis of various loss and lossless compression algorithms. This research work deals with the preprocessing and transformations used to compress a compound image to produce a high compression ratio (CR), less compression time and so on. In the compression process the images are considered for preprocessing and discrete wavelet transform with adaptive particle swarm optimization process. The purpose of this optimization technique is to optimize the wavelet coefficient in Harr wavelet for improving the CR value. In the image compression process, run length coding is used to compress the compound images. Based on this technique, it produces minimum CR and less computation time of compound images.

REGION SPECIFIC WAVELET COMPRESSION FOR 4K SURVEILLANCE IMAGES

For successful transmission of massively sequenced images during 4K surveillance operations large amount of data transfer cost high bandwidth, latency and delay of information transfer. Thus, there lies a need for real-time compression of this image sequences. In this study we present a region specific approach for wavelet based image compression to enable management of huge chunks of information flow by transforming Harr wavelets in hierarchical order.

A new fractional wavelet transform

The fractional Fourier transform (FRFT) is a potent tool to analyze the time-varying signal. However, it fails in locating the fractional Fourier domain (FRFD)-frequency contents which is required in some applications. A novel fractional wavelet transform (FRWT) is proposed to solve this problem. It displays the time and FRFD-frequency information jointly in the time-FRFD-frequency plane. The definition, basic properties, inverse transform and reproducing kernel of the proposed FRWT are considered. It has been shown that an FRWT with proper order corresponds to the classical wavelet transform (WT). The multiresolution analysis (MRA) associated with the developed FRWT, together with the construction of the orthogonal fractional wavelets are also presented. Three applications are discussed: the analysis of signal with time-varying frequency content, the FRFD spectrum estimation of signals that involving noise, and the construction of fractional Harr wavelet. Simulations verify the validity of the proposed FRWT.

An Improved Wavelet PWM Technique with Output Voltage Amplitude Control for Single-phase Inverters

Unlike existing pulse-width modulation (PWM) techniques, such as sinusoidal PWM and random PWM, the wavelet PWM (WPWM) technique based on a Harr wavelet function can achieve a high fundamental component for the output voltage, low total harmonic distortion, and simple digital implementation. However, the original WPWM method lacks output voltage control. Thus, the practical application of the WPWM technique is limited. This study proposes an improved WPWM technique that can regulate output voltage amplitude with the addition of a parameter. The relationship between the additional parameter and the output voltage amplitude is analyzed in detail. Experimental results verify that the improved WPWM exhibits output voltage control in addition to all the merits of the WPWM technique.

An 8.12 μW wavelet denoising chip for PPG detection and portable heart rate monitoring in 0.18 μm CMOS

A low power wavelet denoising chip for photoplethysmography (PPG) detection and portable heart rate monitoring is presented. To eliminate noise and improve detection accuracy, Harr wavelet (HWT) is chosen as the processing tool. An optimized finite impulse response structure is proposed to lower the computational complexity of proposed algorithm, which is benefit for reducing the power consumption of proposed chip. The modulus maxima pair location module is design to accurately locate the PPG peaks. A clock control unit is designed to further reduce the power consumption of the proposed chip. Fabricated with the 0.18 μm N-well CMOS 1P6M technology, the power consumption of proposed chip is only 8.12 μW in 1 V voltage supply. Validated with PPG signals in multiparameter intelligent monitoring in intensive care databases and signals acquired by the wrist photoelectric volume detection front end, the proposed chip can accurately detect PPG signals. The average sensitivity and positive prediction are 99.91% and 100%, respectively.

AUTOMATIC GLUCOSE INSULIN REGULATION SYSTEM-COMPARISON OF EMBEDDED CONTROL DESIGN AND HARR WAVELET METHOD FOR TYPE-1 DIABETES

This article is aimed to inject insulin for type 1 diabetes using pumping the insulin with respect to meal disturbance. We have used the embedded linear parameter varying methodology controller and the final results were compared with the Harr wavelet mathematical values. In this article, there are three things we need to focus. First is the sensor values which is to be monitored from the sensors and next is the lab information. Depending on the patients test details the insulin has to be injected. The comparison of lab details with the Patients current sensor values play a vital role in determining the insulin level for a patient. Finally the feedback has to be obtained with the help of those comparisons and it has to be sent once again as a loop to the controller for later comparison and also for database information. The embedded controller is the key element for updating all the information about the patient and it will control all the parameters of the board. The experimental results shows that the proposed method has much better potential in terms of solution accuracy as 97% and better convergence speed in comparison with Haar Wavelet Model which proved 89.09% of accuracy of the assured values of our experiments.

Application of Wavelet Filter Method in Optical Tweezers Signal Processing

One limitation on the performance of optical tweezers (abbreviated as OT) is the noise inherently present in each setup. Therefore, it is the desire to minimize and possibly eliminate the noise from the OT experiments. In this paper, a filter method based on wavelet analysis is proposed. At first we investigate the properties of OT outputs noise, and introduce the wavelet filtering method in simply. Following, we study on the OTs drift signal using different base: db4 and Haar. And also study on the signal using different filter algorithm: the soft,the hard threshold,and compulsive filter. These main conclusions based on foregoing analysis are reached: more larger the resolving scale is, more perfect the filtering effect is. The soft threshold value filtering effect is better than that of the hard threshold value filtering at the cost of calculation when the threshold value is same. The variance of the compulsive filtering is least when both the wavelet and the resolving scale are same for these filtering methods. For the compulsive filtering with same wavelets, the filtering effect of harr is better than that of db4 and the calculation of the former is fewer. Analysis the dynamic output of OT with different algorithm, it also shows that the effect of filter with the compulsive filtering is better than others. Accordingly, we found that applying the compulsive filtering with the Harr wavelet base and suitable resolving scale to the signal processing of OT outputs signal is helpful for the OT design and construction.

Vehicle Detection Using Probabilistic Fusion

This paper investigates the effectiveness of probabilistic decision-level fusion in the context of vehicle detection from gray-scale images. Specifically, we demonstrate that probabilistic fusion rules often disregard the “physical” meaning of classifier outputs and make assumptions that might not hold in practice. As a result, their performance is seriously affected. To support our argument, we have trained two Support Vector Machine (SVM) classifiers to perform vehicle detection using two types of features: Harr Wavelet and Gabor. To classify an input pattern, the output of the SVM classifier needs to be thresholded first. In this case, each output represents a class. When considering the raw SVM outputs (i.e., without thresholding), however, each output represents a “distance” between the input pattern and the decision boundary. Unfortunately, some popular probabilistic decision-level fusion rules, such as the Sum and Product rules, disregard the physical meaning of the raw SVM outputs. Moreover, they make assumptions about data independence and distribution models which might not hold in practice. Motivated by these observations, we propose a simple but effective decision-level fusion rule which exploits the physical meaning of the SVM outputs and does not make any assumptions about the data. We have evaluated the proposed rule on real data sets, showing that it outperforms traditional probabilistic fusion rules.

Two Types of Haar Wavelet Neural Networks for Nonlinear System Identification

Since wavelet transform uses the multi-scale (or multi-resolution) techniques for time series, wavelet transform is suitable for modeling complex signals. Haar wavelet transform is the most commonly used and the simplest one. The Haar wavelet neural network (HWNN) applies the Harr wavelet transform as active functions. It is easy for HWNN to model a nonlinear system at multiple time scales and sudden transitions. In this paper, two types of HWNN, feedforward and recurrent wavelet neural networks, are used to model discrete-time nonlinear systems, which are in the forms of the NARMAX model and state-space model. We first propose an optimal method to determine the structure of HWNN. Then two stable learning algorithms are given for the shifting and broadening coefficients of the wavelet functions. The stability of the identification procedures is proven.