Farrukh Nagi

Max-pooling convolutional neural networks for vision-based hand gesture recognition

Automatic recognition of gestures using computer vision is important for many real-world applications such as sign language recognition and human-robot interaction (HRI). Our goal is a real-time hand gesture-based HRI interface for mobile robots. We use a state-of-the-art big and deep neural network (NN) combining convolution and max-pooling (MPCNN) for supervised feature learning and classification of hand gestures given by humans to mobile robots using colored gloves. The hand contour is retrieved by color segmentation, then smoothened by morphological image processing which eliminates noisy edges. Our big and deep MPCNN classifies 6 gesture classes with 96% accuracy, nearly three times better than the nearest competitor. Experiments with mobile robots using an ARM 11 533MHz processor achieve real-time gesture recognition performance.

Automated breast profile segmentation for ROI detection using digital mammograms

Mammography is currently the most effective imaging modality used by radiologists for the screening of breast cancer. Finding an accurate, robust and efficient breast profile segmentation technique still remains a challenging problem in digital mammography. Extraction of the breast profile region and the pectoral muscle is an essential pre-processing step in the process of computer-aided detection. Primarily it allows the search for abnormalities to be limited to the region of the breast tissue without undue influence from the background of the mammogram. The presence of pectoral muscle in mammograms biases detection procedures, which recommends removing the pectoral muscle during mammogram pre-processing. In this paper we explore an automated technique for mammogram segmentation. The proposed algorithm uses morphological preprocessing and seeded region growing (SRG) algorithm in order to: (1) remove digitization noises, (2) suppress radiopaque artifacts, (3) separate background region from the breast profile region, and (4) remove the pectoral muscle, for accentuating the breast profile region. To demonstrate the capability of our proposed approach, digital mammograms from two separate sources are tested using Ground Truth (GT) images for evaluation of performance characteristics. Experimental results obtained indicate that the breast regions extracted accurately correspond to the respective GT images.

A computational intelligence scheme for the prediction of the daily peak load

Forecasting of future electricity demand is very important for decision making in power system operation and planning. In recent years, due to privatization and deregulation of the power industry, accurate electricity forecasting has become an important research area for efficient electricity production. This paper presents a time series approach for mid-term load forecasting (MTLF) in order to predict the daily peak load for the next month. The proposed method employs a computational intelligence scheme based on the self-organizing map (SOM) and support vector machine (SVM). According to the similarity degree of the time series load data, SOM is used as a clustering tool to cluster the training data into two subsets, using the Kohonen rule. As a novel machine learning technique, the support vector regression (SVR) is used to fit the testing data based on the clustered subsets, for predicting the daily peak load. Our proposed SOM-SVR load forecasting model is evaluated in MATLAB on the electricity load dataset provided by the Eastern Slovakian Electricity Corporation, which was used in the 2001 European Network on Intelligent Technologies (EUNITE) load forecasting competition. Power load data obtained from (i) Tenaga Nasional Berhad (TNB) for peninsular Malaysia and (ii) PJM for the eastern interconnection grid of the United States of America is used to benchmark the performance of our proposed model. Experimental results obtained indicate that our proposed SOM-SVR technique gives significantly good prediction accuracy for MTLF compared to previously researched findings using the EUNITE, Malaysian and PJM electricity load datasets.

Improving SVM-Based Nontechnical Loss Detection in Power Utility Using the Fuzzy Inference System

This letter extends previous research work in modeling a nontechnical loss (NTL) framework for the detection of fraud and electricity theft in power distribution utilities. Previous work was carried out by using a support vector machine (SVM)-based NTL detection framework resulting in a detection hitrate of 60%. This letter presents the inclusion of human knowledge and expertise into the SVM-based fraud detection model (FDM) with the introduction of a fuzzy inference system (FIS), in the form of fuzzy IF-THEN rules. The FIS acts as a postprocessing scheme for short-listing customer suspects with higher probabilities of fraud activities. With the implementation of this improved SVM-FIS computational intelligence FDM, Tenaga Nasional Berhad Distributions detection hitrate has increased from 60% to 72%, thus proving to be cost effective.

Convolutional Neural Support Vector Machines: Hybrid Visual Pattern Classifiers for Multi-robot Systems

We introduce Convolutional Neural Support Vector Machines (CNSVMs), a combination of two heterogeneous supervised classification techniques, Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs). CNSVMs are trained using a Stochastic Gradient Descent approach, that provides the computational capability of online incremental learning and is robust for typical learning scenarios in which training samples arrive in mini-batches. This is the case for visual learning and recognition in multi-robot systems, where each robot acquires a different image of the same sample. The experimental results indicate that the CNSVM can be successfully applied to visual learning and recognition of hand gestures as well as to measure learning progress.

NTL detection of electricity theft and abnormalities for large power consumers In TNB Malaysia

Electricity consumer dishonesty is a problem faced by all power utilities. Finding efficient measurements for detecting fraudulent electricity consumption has been an active research area in recent years. This paper presents an approach towards detection of Non-technical Losses (NTLs) of Large Power Consumers (LPC) in Tenaga Nasional Berhad (TNB) Malaysia. The main motivation of this study is to assist Tenaga Nasional Berhad (TNB) Sdn. Bhd. in Malaysia to reduce its NTLs in the LPC distribution sector. Remote meters installed at premises of LPC customers transmit power consumption data including remote meter events wirelessly to TNB Metering Services Sdn. Bhd. The remote meter reading (RMR) consumption data for TNB LPC customers is recorded based on half-hourly intervals. The technique proposed in this paper correlates the half-hourly RMR consumption data with abnormal meter events. The correlated data provides information regarding consumption characteristics i.e. load profiles of LPC customers, which helps to expose abnormal consumption behavior that is known to be highly correlated with NTL activities and electricity theft. Pilot testing results obtained from TNB Distribution (TNBD) Sdn. Bhd. for onsite inspection of LPC customers in peninsular Malaysia indicate the proposed NTL detection technique is effective with a 55% detection hitrate. With the implementation of this intelligent system, NTL activities of LPC customers in TNB Malaysia will reduce significantly.

Fuzzy bang-bang relay controller for satellite attitude control system

Two level bang-bang controllers are generally used in conjunction with the thrust reaction actuator for spacecraft/satellite attitude control. These controllers are fast acting and dispense time dependent; full or no thrust-power to control the satellite attitude in minimum time. A minimum time-fuel attitude control system extends the life of a satellite and is the main focus of this paper. Fuzzy controllers are favored for satellite control due to their simplicity and good performance in terms of fuel saving, absorbing non-linearities and uncertainties of the plant. A fuzzy controller requires a soft fuzzy engine, and a hardware relay to accomplish bang-bang control action. The work in this paper describes a new type of fuzzy controller in which the hardware relay action is configured in the soft fuzzy engine. The new configuration provides fuzzy decision-making flexibility at the inputs with relay like two-level bang-bang output. The new fuzzy controller is simulated on a three-axis satellite attitude control platform and compared with conventional a fuzzy controller, sliding mode controller and linear quadratic regulator. The result shows that the proposed controller has minimum-time response compared to other controllers. Inherent chattering associated with a two-level bang-bang controller produces undesirable low amplitude frequency limit cycles. The chattering can be easily stopped in the proposed fuzzy bang-bang relay controller, hence adding multi-functionality to its simple design.

DSP based overcurrent relay using fuzzy bang–bang controller

Abstract Overcurrent relays are very important protection component that require high reliability to maintain high security in power systems. With the new numerical relay technology using digital signal processor (DSP), it is possible to improve the performance of the relay significantly. However, application of DSP in numerical overcurrent relays is limited especially in coordination among the group of relays. The relay must work proficiently to coordinate with the networks in order to avoid mal-operation. Therefore, in this paper, an implementation of overcurrent relay with improved coordination on a DSP, TMS320F2812 is described. The fuzzy bang–bang controller is used as the control strategy for the relay to provide efficient control for overcurrent protection. The performance evaluation of the proposed system is based on steady state analysis, transient state analysis, coordination and lastly the execution time of the DSP. The results obtained using this new proposed controller is very promising.

Fuzzy bang-bang relay control of a single-axis active magnetic bearing system

Abstract Active magnetic bearings (AMB) are presently being utilized in various classes of rotating machinery. Although the rotor-AMB systems are open loop unstable, they are easily stabilized using feedback control schemes of which the PID controller is the most commonly used. The PID controller is however only effective at the vicinity of the rotor’s equilibrium position where the dynamics of the rotor-AMB system is linearized. Significant deviation of the rotor’s motion from this equilibrium position may occur due to large imbalance forces. In this situation, the nonlinearity in AMBs, which arises from the relationship between the electromagnetic force, coil current and air gap, may render the PID controller ineffective. For the control of nonlinear systems, artificial intelligence techniques such as fuzzy and hybrid techniques are effective. In this paper, a new fuzzy controller is proposed for the control of a single-axis AMB system. This controller is based on the bang–bang scheme, which is an old but effective technique to control nonlinear systems in optimal time. The performance of the proposed integrated fuzzy bang–bang relay controller (FBBRC) was found to be superior to that of the optimized PD controller and the conventional fuzzy logic controller. Comparison of the FBBRC with the fuzzy logic controller cascaded with a hard limiter (FBBC) relay revealed almost equal performance. High frequency chattering was however observed in the steady-state response of the FBBC. Such chattering is known to cause instability and distortion in the amplifiers that are used to supply current to the magnetic bearing actuators.

DSP based fuzzy and conventional overcurrent relay controller comparisons

Abstract Fuzzy logic control uses linguistic approach to solve complicated rules and ambiguous systems. This control strategy can be used to improve the overall performance of an overcurrent relay for power system protection compared to conventional relay. It is essential for a relay to work efficiently to trip the circuit breakers in the presence of faults and at the same time proficient to coordinate well with the networks to avoid mal-operation. There are two different types of fuzzy logic control strategies proposed for the relay, the Fuzzy Logic Controller (FLC) and Fuzzy Bang-Bang Controller (FBBC). The FBBC is the same as the conventional FLC except that the defuzzification method uses largest of maxima (LOM). Comparisons between the fuzzy controllers and conventional relay are based on IEC 255-3 standard. These relays are implemented on a DSP TMS320F2812 and their performance is evaluated which is based on operation time, DSP’s execution time and grading margin. The results obtained show a significant performance improvement compared to conventional relay.