Hill-climbing Search Research Articles

Research on the Control Strategy of Maximum Energy Utilization and Disordered Disturbance Suppression in Distributed Generation System

The ultimate goal of distributed renewable energy control is to minimize disorder disturbance to the power grid while ensuring maximum energy utilization. Taking wind power as an example, a coupling system using flywheel energy storage system to balance the output variables of wind power system was proposed while using hill climb searching (HCS) to maximize the utilization of wind energy. These methods can obtain stable DC output voltage. In the case of wind speed variation, as a limited capacity energy pool structure, flywheel energy storage system can ensure the maximum utilization of wind energy. Based on fully utilization of wind energy resources, the rapid and comprehensive compensation of wind power output can effectively smooth the output voltage of grid-connected wind power system, thus achieving the purpose of improving the power quality of grid-connected wind power system.

Neural Architecture Search for Skin Lesion Classification

Deep neural networks have achieved great success in many domains. However, successful deployment of such systems is determined by proper manual selection of the neural architecture. This is a tedious and time-consuming process that requires expert knowledge. Different tasks need very different architectures to obtain satisfactory results. The group of methods called the neural architecture search (NAS) helps to find effective architecture in an automated manner. In this paper, we present the use of an architecture search framework to solve the medical task of malignant melanoma detection. Unlike many other methods tested on benchmark datasets, we tested it on practical problem, which differs greatly in terms of difficulty in distinguishing between classes, resolution of images, data balance within the classes, and the number of data available. In order to find a suitable network structure, the hill-climbing search strategy was employed along with network morphism operations to explore the search space. The network morphism operations allow for incremental increases in the network size with the use of the previously trained network. This kind of knowledge reusing allows significantly reducing the computational cost. The proposed approach produces structures that achieve similar results to those provided by manually designed structures, at the same time making use of almost 20 times fewer parameters. What is more, the search process lasts on average only 18h on single GPU.

Maximizing Achievable Transmission Time in Cognitive Radio Networks Under Sensor-Aided Crowdsourced Spectrum Sensing

Abstract Spectrum-sensing as a service has been proposed and studied by many researchers over the past decade as a promising approach to support the viability of cognitive radio networks (CRNs). A spectrum-sensing service provider (SSP) provides information about spectrum occupancy to its clients that is generally more accurate than what clients can learn on their own. Two approaches are used by SSPs in their operation, the dedicated sensing infrastructure approach (sensor-aided CRN) and the crowdsensing approach. In this work, we assume a hybrid model where a dedicated sensing infrastructure is used along with crowdsensing. We study the tradeoff between sensing time paid by cognitive users to the SSP and their achievable transmission time. Our objective is to maximize the minimum achievable transmission time for any cognitive user in the network by carefully selecting the channels to be used. Two algorithms are proposed, one is based on the hill-climbing search algorithm (abbreviated HCA) and the other is a less optimal but faster greedy selection algorithm (abbreviated GSA). Results show that both HCA and GSA are within 3% of the optimal solution. Results also confirm that GSA is faster than HCA, while HCA outperforms GSA.

Sliding mode extremum seeking control based on improved invasive weed optimization for MPPT in wind energy conversion system

The sliding mode extremum seeking control (SMESC) could track the maximum power point (MPP) of wind energy conversion system (WECS) without wind speed or wind turbine parameters. Inappropriate SMESC parameters would cause steady-state oscillation and increase tracking time. This paper proposed an improved invasive weed optimization (IIWO) to optimize the SMESC parameters. The algorithm developed a new stochastic reproductive strategy to enhance its robustness and simplify the coding. Meanwhile, IIWO optimized double parameters coordinately to replace traditional parameter setting methods of SMESC, which could make the parameters meet the different requirements simultaneously for high efficiency. Simulation results showed that proposed IIWO-SMESC method yielded a better transient response, steady-state stability, and robustness than traditional hill-climbing search (HCS) and SMESC method.

IIWO Based Sliding Mode Extremum Seeking Control for MPPT in Wind Energy Conversion System

Abstract The sliding mode extremum seeking control (SMESC) could track the maximum power point (MPP) of wind energy conversion system (WECS) without wind speed or wind turbine parameters. Inappropriate SMESC parameters would cause steady-state oscillation and increase tracking time. This paper proposed an improved invasive weed optimization (IIWO) to optimize the SMESC parameters. The algorithm developed a new stochastic reproductive strategy to enhance its robustness and simplify the coding. Meanwhile, IIWO optimized double parameters coordinately to replace traditional parameter setting methods of SMESC, which could make the parameters meet the different requirements simultaneously for high efficiency. Simulation results showed that proposed IIWO-SMESC method yielded a better transient response, steady-state stability, and robustness than conventional hill-climbing search (HCS) and SMESC method.

Matching for navigation map building for automated guided robot based on laser navigation without a reflector

PurposeLaser navigation without a reflector does not require setup of reflector markers at the scene and thus has the advantages of free path setting and flexible change. This technology has attracted wide attention in recent years and shows great potential in the field of automatic logistics, including map building and locating in real-time according to the environment. This paper aims to focus on the application of feature matching for map building.Design/methodology/approachFirst, an improved linear binary relation algorithm was proposed to calculate the local similarity of the feature line segments, and the matching degree matrix of feature line segments between two adjacent maps was established. Further, rough matching for the two maps was performed, and both the initial rotation matrix and the translation vector for the adjacent map matching were obtained. Then, to improve the rotation matrix, a region search optimization algorithm was proposed, which took the initial rotation matrix as the starting point and searched gradually along a lower error-of-objective function until the error sequence was nonmonotonic. Finally, the random-walk method was proposed to optimize the translation vector by iterating until the error-objective function reached the minimum.FindingsThe experimental results show that the final matching error was controlled within 10 mm after both rotation and translation optimization. Also, the algorithm of map matching and optimization proposed in this paper can realize accurately the feature matching of a laser navigation map and basically meet the real-time navigation and positioning requirements for an automated-guided robot.Originality/valueA linear binary relation algorithm was proposed, and the local similarity between line segments is calculated on the basis of the binary relation. The hill-climbing region search algorithm and the random-walk algorithm were proposed to optimize the rotation matrix and the translation vector, respectively. This algorithm has been applied to industrial production.

Bnclassify: Learning Bayesian Network Classifiers

The bnclassify package provides state-of-the art algorithms for learning Bayesian network classifiers from data. For structure learning it provides variants of the greedy hill-climbing search, a well-known adaptation of the Chow-Liu algorithm and averaged one-dependence estimators. It provides Bayesian and maximum likelihood parameter estimation, as well as three naive-Bayes-specific methods based on discriminative score optimization and Bayesian model averaging. The implementation is efficient enough to allow for time-consuming discriminative scores on medium-sized data sets. bnclassify provides utilities for model evaluation, such as cross-validated accuracy and penalized log-likelihood scores, and analysis of the underlying networks, including network plotting via the Rgraphviz package. It is extensively tested, with over 200 automated tests that give a code coverage of 94%. Here we present the main functionalities, illustrate them with a number of data sets, and comment on related software.

Adaptive Step Size Based Hill-Climb Search Algorithm for MPPT Control of DFIG-WECS With Reduced Power Fluctuation and Improved Tracking Performance

One of the major challenges in harnessing wind energy is to extract maximum power from intermittent generation of wind farms as wind power generation strongly depends on wind speed variation. Among different maximum power point tracking (MPPT) algorithms, traditionally, hill climb search (HCS) method is widely used because of its simple implementation and turbine parameter-independence. However, the conventional HCS algorithm has some drawbacks such as power fluctuation and speed-efficiency tradeoff. Similarly constant tip-speed ratio (TSR) MPPT control method requires precise knowledge of optimal TSR of the turbine. Therefore, in this paper a new adaptive step size based HCS controller is proposed to mitigate the deficiencies of other techniques by incorporating wind speed measurement. The function-based adaptive control scheme evaluates the step size by the variation of the wind speed and extracted power range. The proposed variable step size based HCS-MPPT system is simulated at variable wind speed. A prototype system is also built with a low-power doubly-fed induction generator (DFIG). It is found from the results that the proposed controller reduces the steady-state power fluctuation of the DFIG based wind energy conversion system significantly compared to the conventional HCS-MPPT controller and constant TSR based MPPT controller at variable wind speeds.

Computational Synthesis of Large Deformation Compliant Mechanisms Undergoing Self and Mutual Contact

Topologies of large deformation contact-aided compliant mechanisms (CCMs), with self and mutual contact, exemplified via path generation applications, are designed using the continuum synthesis approach. Design domain is parameterized using honeycomb tessellation. Assignment of material to each cell, and generation of rigid contact surfaces, are accomplished via suitably sizing and positioning negative circular masks using the stochastic hill-climber search. To facilitate contact analysis, boundary smoothing is implemented. Mean value coordinates are employed to compute shape functions, as many regular hexagonal cells get degenerated into irregular, concave polygons as a consequence of boundary smoothing. Both geometric and material nonlinearities are considered. The augmented Lagrange multiplier method with a formulated active set strategy is employed to incorporate both self and mutual contact. Synthesized contact-aided compliant continua trace paths with single, and importantly, multiple kinks and experience multiple contact interactions pertaining to both self and mutual contact modes.

Enhanced radial fuzzy wavelet neural network with sliding mode control for a switched reluctance wind turbine distributed generation system

ABSTRACTIn this article, an Enhanced Radial Fuzzy Wavelet Neural Network with Sliding Mode (RFWNNSM) and hill-climb searching (HCS) maximum power point tracking (MPPT) strategy are proposed for a switched reluctance generator (SRG) in a variable-speed wind energy conversion system (WECS). The implemented MPPT control is an HCS algorithm that does not require the knowledge of turbine or generator characteristics. However, this study proposes a simple optimal DC-Link voltage search type HCS control, and therefore a faster convergence to MPPT is achieved. A high-performance online training Enhanced Radial Fuzzy Wavelet Neural Network (RFWNN) using a back-propagation learning algorithm with a sliding mode regulating controller is designed for an SRG. The MPPT strategy locates the system operation points along the maximum power curves based on the DC-Link voltage of the inverter, thus avoiding the need to detect the generator speed.

Towards transfer synchronization of regularity-based bus operations with sequential hill-climbing

In this work we model and discuss how we can achieve coordination between different bus service lines. Key problem challenges are (a) the multiple conflicting priorities (on one hand the improvement of bus service regularity and on the other hand the reduction of passenger transfer waiting times) and (b) the computational complexity for re-scheduling the dispatching times of bus trips for meeting the conflicting priorities. Initially, a model for reducing the waiting times at bus transfer stations while also improving the operations of regularity-based bus services subject to operational constraints is introduced. Conflicting priorities are handled with the introduction of weight factors that allow bus operators to decide the trade-off between improvement of regularity-based operations and reduction of passenger waiting times at transfer stations. After that, an exterior point penalty function is introduced for handling operational constraints and a sequential hill-climbing search strategy is applied for converging to an approximate optimal solution. For our case study, we utilize general transit feed specification data from two regularity-based bus services in central Stockholm that intersect in five transfer stations. Experimental tests showcase a 13% potential waiting time improvement at transfer stations while sacrificing only 2.8% of service regularity and satisfying all operational constraints.

Metaheuristic Approach for Distributed Trajectory Planning for European Functional Airspace Blocks

The functional airspace block (FAB) concept is adopted by European airspace to allow cooperation between airspace users to manage the air traffic flow, while ensuring efficiency, safety, and fairness without the constraints of geographical boundaries. This integration of airspaces allows for flexibility in airspace management and aircraft trajectory planning. This paper proposes a distributed air traffic flow management model to address four-dimensional trajectory planning over the European FAB. The proposed method is based on a metaheuristic approach that uses a hybrid algorithm of simulated annealing and hill-climbing local search to separate a given set of aircraft trajectories in space and time domain (termed as flight interaction), by allocating an alternative flight plan (route and departure time) to each flight. An innovative data structure, termed as FAB–flight interaction matrix, captures the flight interaction information between and within FABs. The proposed distributed model is implemented and tested with two air traffic data sets comprising 4000 flights (3 h traffic) and 26,000 flights (one full day traffic data over the European airspace). The performance of the model is then compared with a centralized air traffic flow management model on scalability and interaction minimization. Results indicates that, though both approaches were able to achieve interaction-free trajectory planning within computational time acceptable for the operational context, the distributed model converges faster to an interaction-free solution as traffic size increases, which shows the viability of the distributed model for effective FAB implementation.

Component Sizing of Parallel Hybrid Electric Vehicle Using Optimal Search Algorithm

In designing a parallel hybrid electric vehicle, it is essential to select the optimal capacity of power sources and the optimal gear ratio of the torque coupler. The capacity of the power sources and the gear ratio of the torque coupler should be optimized simultaneously. However, since this process is excessively time-consuming, previous studies have selected the gear ratio of the torque coupler and then selected the capacity of power source. However, this approach cannot guarantee global optimization. In this paper, a feasible region is defined to satisfy the required performance of vehicle such as maximum speed, hill-climbing. and feasible points are selected inside the feasible region. In the conventional technique, the global optimal solution is obtained by simulating all feasible points. In the optimization technique, optimal points are simulated within the feasible region using several optimal search algorithms, such as the golden section search algorithm and the hillclimbing search algorithm. And using these algorithms, the number of simulations is reduced and the capacity of the power source and the gear ratio of the torque coupler are optimized simultaneously. Finally, the validity of the component sizing results is verified by comparing the global optimal solution obtained by applying the conventional technique with the solution obtained by applying the proposed optimization technique.

Analysis and Comparison of Bayesian Methods for Measurement Uncertainty Evaluation

Based on the Bayesian principle, the modern uncertainty evaluation methods can fully integrate prior and current sample information, determine the prior distribution according to historical information, and deduce the posterior distribution by integrating prior distribution and the current sample data with the Bayesian model. As such, it is possible to evaluate uncertainty, updating in real time the uncertainty of the measuring instrument according to regular measurement, and timely reflect the latest information on the accuracy of the measurement system. Based on the Bayesian information fusion and statistical inference principle, the model of uncertainty evaluation is established. The maximum entropy principle and the hill-climbing search optimization algorithm are introduced to determine the prior distribution probability density function and the sample information likelihood function. The probability density function of posterior distribution is obtained by the Bayesian formula to achieve the optimization estimation of uncertainty. Three methods of measurement uncertainty evaluation based on Bayesian analysis are introduced: the noninformative prior, the conjugate prior, and the maximum entropy prior distribution. The advantages and limitations of each method are discussed.

Comparative study of different MPPT methods for wind energy conversion system

In this paper we present a comparative study of different MPPT methods for a wind energy system based on permanent magnet synchronous generator (PMSG). These techniques are: Hill Climbing Search (HCS), Optimal Torque Control (OTC), Power Signal Feedback (PSF) and Fuzzy Logic control (FLC). These methods are developed with the aim of comparing their performance on the wind system energy efficiency, according to the response time, the maximum power to be achieved and the system behaviour during the optimal power point tracking MPP. And for a better analysis study, we simulated the system under variable wind speed conditions. The results show that intelligent control based on fuzzy logic is more efficient and presents better performance compared to other MPPT algorithms.

On defining a unique phylogenetic tree with homoplastic characters

This paper discusses the problem of whether creating a matrix with all the character state combinations that have a fixed number of steps (or extra steps) on a given tree T, produces the same tree T when analyzed with maximum parsimony or maximum likelihood. Exhaustive enumeration of cases up to 20 taxa for binary characters, and up to 12 taxa for 4-state characters, shows that the same tree is recovered (as unique most likely or most parsimonious tree) as long as the number of extra steps is within 1/4 of the number of taxa. This dependence, 1/4 of the number of taxa, is discussed with a general argumentation, in terms of the spread of the character changes on the tree used to select character state distributions. The present finding allows creating matrices which have as much homoplasy as possible for the most parsimonious or likely tree to be predictable, and examination of these matrices with hill-climbing search algorithms provides additional evidence on the (lack of a) necessary relationship between homoplasy and the ability of search methods to find optimal trees.

Artificial neural controller for torque ripple control and maximum power extraction for wind system driven by switched reluctance generator

Abstract Torque ripples and maximum power tracking form the most important problems facing Wind Energy Conversion System (WECS). This paper presents modified toque ripple minimization algorithm of four phases 8/6 poles Switched Reluctance Generator (SRG) using Artificial Neural Network (ANN) control. This control algorithm is based on optimum profiling of the generator phase currents during overlapping periods. The paper, also, introduces new Maximum Power Point Tracking (MPPT) for SRG by modifying the classical Hill Climb Searching (HCL) technique using ANN which is designed to emulate the PI controller for closed loop operation under different wind speeds. Moreover multi level diode clamped inverter is used for connecting SRG to the grid in order to reduce Total Harmonic Distortion (THD) to about 1.45% and reduce filter size. The simulation results show good agreement and support the feasibility of the proposed torque ripple minimization and MPPT techniques.

Modeling and optimization of polymer enhanced ultrafiltration using hybrid neural-genetic algorithm based evolutionary approach

A stochastic genetic algorithm (GA) based strategy along with artificial neural network (ANN) was applied to optimize the retention of reactive red 120 (RR 120) dye from its aqueous solutions by way of polymer (polyethyleneimine (PEI)) enhanced ultrafiltration (PEUF). The optimal feed forward back propagation ANN (4-10-1) model network, trained initially through Levenberg–Marquardt (LM) algorithm, was suitably manoeuvred by the GA approach to predict the membrane performance index (PFI) response, evaluated as the product of dye rejection and permeation flux, for a randomly generated population of chromosomes. Each chromosome was constituted by four principal genes, namely, cross-flow rate, transmembrane pressure, polymer to dye ratio, and pH. The local exploitation capacity of the canonical GA was enhanced further by combining hill-climbing (HC) local search with the optimization levels of standard GA. The near-optimal and economically feasible factor levels were predicted by the hybrid ANN-GA-HC strategy, keeping PFI maximization and the constrained PEUF process dynamics in perspective; the optimal process factor settings experimentally yielded a pragmatic PFI of 143.8L/m2h, corresponding to high (99.9%) dye rejection, and a satisfactory permeation flux (144L/m2h).

A 0.45-to-1.2-V Fully Digital Low-Dropout Voltage Regulator With Fast-Transient Controller for Near/Subthreshold Circuits

A low quiescent current digital low-dropout (D-LDO) voltage regulator with fast-transient response time is proposed for self-powered wireless sensor applications operating at near/subthreshold supply voltage. The D-LDO regulator incorporates both hill-climbing and binary search algorithms (HCBS) in the control logic, thus leveraging on each other's strengths to minimize the output voltage's ripple and the quiescent current during the steady-state period as well as output voltage's spike and response time during the transition period. Additional features such as hysteresis mode control and freeze mode control are incorporated into the system to improve the performance of the D-LDO regulator. A dynamic comparator is proposed for the near/subthreshold supply voltage operation, which minimizes the voltage error and improves the maximum operating frequency. Fabricated in 130-nm CMOS technology, the D-LDO regulator regulates the output voltage ${V}_{{\rm OUT}}$ from 350 to 1150 mV, while the input supply voltage ${V}_{\text{IN}}$ ranges from 450 to 1200 mV. At a ${V}_{{\rm OUT}}$ of 450 mV, ${V}_{{\rm IN}}$ of 500 mV and an operating frequency of 10 MHz, the regulator delivers 1500- ${\mu }$ A load current with $I_{\text{QUIESCENT}}$ of 8.9 ${\mu }$ A and a transient response time of 1.6 ${\mu }$ s. The maximum current and power efficiencies reach 99.9% and 89.9%, respectively. The measured line regulation and load regulation are 1.6 and 0.6 mV/mA, respectively.

Design of a Unified Power Controller for Variable-Speed Fixed-Pitch Wind Energy Conversion System

A unified power controller for variable-speed fixed-pitch wind energy conversion system (WECS) is designed covering the whole range of wind speed in this paper. The proposed controller is composed of a modified maximum power point tracking (MPPT) controller in low wind speed and a new constant speed and constant power controller in high wind speed. For the former, a combination of modified hill climb searching (HCS) and power signal feedback (PSF) MPPT algorithms is used. The modified HCS method is activated to search for the maximum power point (MPP) first, which is followed by the PSF method once one MPP is found. By using this controller, not only the a priori knowledge of the aerodynamic characteristics of turbine blades is avoided, but also low torque/power ripple is achieved; for the latter, a new auxiliary passive stall control method is proposed. It temporarily increases the output power to force the turbine to operate in deep stall regime, thus to decrease the captured power of the turbine. The proposed controller is implemented on a digital signal processor. The validity of the proposed method is verified by experimental results done on a 10-kW WECS.