Hierarchical Search Strategy Research Articles

Improved nonlinear active disturbance rejection control based on hierarchical expected improvement for high-speed elevator horizontal vibration suppression

In this paper, an improved NADR control method based on hierarchical expected improvement (HEI) optimization is studied to suppress the horizontal vibration of high-speed elevator car under complex coupled disturbances to improve high-speed elevator ride comfort. First, a nonlinear 8-DOF dynamic model for the horizontal vibration of the high-speed elevator car is constructed. Then, a 4-level parallel NADR main control system based on the fal-function is designed to observe the total disturbance in real time, which can be quickly compensated by fuzzy sliding mode control. Meanwhile, the adaptive law is used to adjust fuzzy rules and sliding mode switching term coefficients to ensure the stability and adaptability of the control system. To address the problem of multiple bandwidth parameters and high sensitivity of the observer in the NADR control system, the comprehensive RMS experimental data fusion model based on hierarchical Gaussian process is constructed. With the goal of achieving the optimal value of comprehensive RMS, a hierarchical expectation improvement search strategy that balances exploratory and developmental aspects is studied to improve the optimization accuracy of bandwidth parameters. Finally, the results show that the proposed method can reduce the horizontal vibration acceleration and tilt angle acceleration of the car system by more than 68%.

An enhanced memetic algorithm with hierarchical heuristic neighborhood search for type-2 green fuzzy flexible job shop scheduling

The green flexible job shop has received increasing attention due to the development of modern industry and the improvement of environmental protection awareness. Meanwhile, in customized manufacturing, the processing time cannot be exactly predicted. Therefore, it can be predicted by the type-2 fuzzy system, which can effectively characterize the uncertainty of processing time. This work aims to solve the green type-2 fuzzy flexible job shop problem (T2FJSP), considering minimizing total energy consumption (TEC) and the makespan simultaneously. Memetic algorithms are used for the T2FJSP, however, most of them do not allocate local search opportunities based on the quality of solutions, resulting in the waste of computing resources. In this paper, an enhanced memetic algorithm with hierarchical heuristic neighborhood search (EMAH) is proposed to overcome the shortcoming. EMAH has the following features: (i) a cooperative heuristic initialization is designed by combining six heuristics to generate high-quality solutions; (ii) a hierarchical heuristic neighborhood search strategy is proposed to better enhance the exploitation by allocating computing resources based on the quality of the solutions; and (iii) an energy-saving strategy is designed to decrease the makespan and TEC. The effectiveness of EMAH is verified by comparing it with the state-of-the-art algorithms in a T2FJSP benchmark. The results indicate the superiority of EMAH in solving the T2FJSP.

Novel hierarchical strategies for SPH-centric algorithms on GPGPU

This research further extends an existing state-of-the-art GPU framework of SPH simulation for better performance without any compromise of numerical accuracy and sacrifice of simulation detail. Towards this grand goal, we devise three new strategies. First, we design a novel hierarchical grid to decompose the simulation space, where we can locate particles in a more refined unit space. With this organization, particle distribution coherence in global memory is improved. As a result, global memory operations can be more efficient. Second, based on our well designed hierarchical grid, we propose a hierarchical neighbor search strategy for catering to the heterogeneous distribution of particles in cells, so we can search a neighbor particle in different-level grids. In a cell with dense particles, we can search particles in a higher-level grid whose unit space is more refined, which means we can narrow the search space for decreasing the access of false neighbor particles. In contrast, in a cell with sparse particles, we search particles in a lower-level grid whose unit space is relatively large, which means we can decrease the loop iterations when we search the entire neighbor cell. In this way, we can avoid the unnecessary overhead of loop iterations. After designing a reasonable neighbor search strategy, an efficient thread cooperation strategy can further improve the performance of our framework by realizing more potentials of GPGPU. The existing state-of-the-art method is concentrating on task assignment strategy for taking the full advantage of shared memory. The method decomposes particles in a cell into several tasks and then assigns a cooperation thread array (CTA) to each task. However, this method has not fully considered the uniformity of tasks belonging to the same cell, as these tasks always share a lot of particles with the same neighborhood. Finally, we propose a hierarchical task assignment strategy by merging such successive tasks into a larger task, which means we only load the same neighbor particles once and the corresponding CTAs are arranged to work together to handle the new task. Our method can greatly reduce the overload of neighbor particles and the overhead of neighbor search iterations. Through our comprehensive tests validated in practice, our work can exhibit 1.73× speedup when compared with other state-of-the-art GPGPU frameworks for SPH simulation.

Design of Self-Shielded Uniform Magnetic Field Coil via Modified Pigeon-Inspired Optimization in Miniature Atomic Sensors

This article presents a modified pigeon-inspired optimization method for designing self-shielded uniform magnetic field coils that offers an innovation in miniature atomic sensors. The proposed coil system consists of circular coil pairs distributed on two coaxial cylinder surfaces, which can provide a highly uniform magnetic field at the coil center and rapidly attenuate the external field. This design is transformed into a nonlinear objective optimization problem with constraints, so that the coil geometric parameters can be obtained by the proposed algorithm, instead of solving closed-form equations by pure numerical method. Parallel competition and hierarchical search strategy are firstly introduced to accelerate the convergence speed and avoid falling into the local optimal value. Comparative experiments of several algorithms demonstrate the accuracy, effectiveness, convergence stability of the proposed method. Compared with traditional design methods, this method can automatically find the optimal parameters of the coil according to any constraints. Theoretical contrast shows that compared with the Lee-Whiting coil, the uniformity of proposed coil is improved by about 2 to 3 orders of magnitude and has a significant self-shielded effect. This work is beneficial to the miniaturization of atomic sensors.

Spatial cellular robot in orbital truss collision-free path planning

Abstract. Aiming at the problem of moving path planning of a cellular robot on trusses in space station, a triangular prism truss is taken as the research object, and an optimized ant colony algorithm incorporating a gravitational search algorithm is proposed. The innovative use of the hierarchical search strategy which limits the exploration area, the use of gravity search algorithm to get the optimal solution of truss nodes, and further transform it into the initial value of pheromone in ant colony algorithm, can effectively prevent the algorithm from falling into the local optimal solution in the early stage, and make the optimization algorithm have a faster convergence speed. This paper proposes a heuristic function including the angle between the targets, which can effectively avoid blind search in the early stage and improve the ability of path search. The simulation results show that the path and planning time of the cellular robot can be effectively reduced when choosing truss path.

An Improved Similarity Algorithm Based On Deep Hash and Code Bit Independence

Deep convolutional neural networks have been widely used in image retrieval because of their powerful feature representation capabilities. Due to the high efficiency of hash space, many image retrieval algorithms based on deep hashing emerge. Aiming at the high correlation between the codes bits of hash code, this paper proposes a new end-to-end trainable deep hash algorithm to implement the feature grouping so that each code bit can express its unique partial image information. We add slice layers to reduce the repetitive expression of information and add the pairing loss to expand the difference between different categories of images and improve the recognition ability of the model for different categories. After training, we use the hierarchical search strategy to test the retrieval ability of the classification model. The mean average precision and recall of our algorithm can reach 87.5% and 92% respectively, which achieves great result in the related field and has important guiding significance for the future research of hash algorithm.

Passenger Flow Pushing Assignment Method for an Urban Rail Network Based on Hierarchical Path and Line Decomposition

: For urban rail transit, an environmentally-friendly transportation mode, reasonable passenger flow assignment is the basis of train planning and passenger control, which is conducive to the sustainability of finance, operation and production. With the continuous expansion of the scale of urban rail networks, passenger travel path decision-making tends to be complex, which puts forward higher requirements of networked transportation organization. Based on undirected graphs and the idea of the recursive divide-and-conquer algorithm, this paper proposes a hierarchical effective path search method made up of a three-layer path generation strategy, which consists of deep search line paths, key station paths composed of origin–destination (O-D) nodes and transfer stations, and the station sequence path between the key stations. It can effectively simplify the path search and eliminate obvious unreasonable paths. Comparing the existing research results based on the classical polynomial Logit model, a practical Improved C-Logit multi-path passenger flow assignment model is proposed to calculate the selection ratio of each path in the set of effective paths. Combining the hierarchical path search strategy, the O-D pairs of passenger flow are divided into local-line and cross-line situations. The time-varying cross-line passenger flow is decomposed into a series of passenger sections along the key station paths. A passenger flow pushing assignment algorithm based on line decomposition is designed, which satisfies the dynamic, time-varying and continuous characteristics. The validation of Guangzhou Metro’s actual line network and time-varying O-D passenger demand in 2019 shows that the spatio-temporal distribution results of the passenger pushing assignment have a high degree of coincidence with the actual statistical data.

Robust template matching with angle location using dynamic feature pairs updating

This work proposes a novel method for template matching in the wild. Different from the previous methods to search the matching position, our method obtains further ability on the angle location by dynamically updating corresponding feature pairs and rigid transformation parameters, which result in mutual enhancement of both feature extraction and template location. We propose a robust objective function with a valid feature selection for template matching against to noise disturbance, background changing, object deformation and partial occlusion. A hierarchical search strategy is used by adjusting the size of feature patch to improve the matching effectiveness. In addition, we extend the proposed method to match image sequences. It is beneficial to propagate reasonable feature pairs to a sequential object as initialization, recalling a stable tracking result. Extensive experiments are tested on the public database with challenging images and sequences. Experimental results demonstrate the merits of the proposed method compared to some state-of-the-art matching methods.

Hierarchical search strategy for the efficient detection of gravitational waves from nonprecessing coalescing compact binaries with aligned-spins

In the first two years of Gravitational Wave (GW) Astronomy, half a dozen compact binary coalescences (CBCs) have been detected. As the sensitivities and bandwidths of the detectors improve and new detectors join the network, many more sources are expected to be detected. The goal will not only be to find as many sources as possible in the data but to understand the dynamics of the sources much more precisely. Standard searches are currently restricted to a smaller parameter space which assumes aligned spins. Construction of a larger and denser parameter space, and optimising the resultant increase in false alarms, pose a serious computational challenge. We present here a two-stage hierarchical strategy to search for CBCs in data from a network of detectors and demonstrate the computational advantage in real life scenario by introducing it in the standard {\tt PyCBC} pipeline with the usual restricted parameter space. With this strategy, in simulated data containing stationary Gaussian noise, we obtain a computational gain of $\sim 20$ over the flat search. In real data, we expect the computational gain up to a factor of few. This saving in the computational effort will, in turn, allow us to search for precessing binaries. Freeing up computation time for the regular analyses will provide more options to search for sources of different kinds and to fulfil the never-ending urge for extracting more science out of the data with limited resources.

Fabric Image Retrieval System Using Hierarchical Search Based on Deep Convolutional Neural Network

Fabric image retrieval is a meaningful issue, due to its potential values in many areas such as textile product design, e-commerce, and inventory management. Meanwhile, it is challenging because of the diversity of fabric appearance. Encourage by the recent breakthrough in the deep convolutional neural network (CNN), a deep learning framework is applied for fabric image retrieval. The idea of the proposed framework is that the binary code and feature for representing the image can be learning by a deep CNN when the data labels are available. The proposed framework employs a hierarchical search strategy that includes coarse-level retrieval and fine-level retrieval. Otherwise, a large-scale wool fabric image retrieval dataset named WFID with about 20 000 images are built to validate the proposed framework. The longitudinal comparison experiments for self-parameter optimization and horizontal comparison experiments for verifying the superiority of the algorithm are performed on this data set. The comparison experimental results indicate the superiority of the proposed framework.

Automated vision positioning system for dicing semiconductor chips using improved template matching method

This study proposes an automated vision positioning system to realize high-efficient and high-precision positioning and dicing of semiconductor chips in an automatic dicing saw. In this method, image pyramid construction was established to improve the searching speed of feature images by using the pyramid hierarchical search strategy. Hough transformation was used to obtain the approximate angle of the feature images of the semiconductor chips. The improved template matching approach based on the initial angle was proposed to rapidly calculate rotation angle and feature position. Polynomial fitting was adopted to achieve sub-pixel positioning accuracy. Experimental results showed that the proposed algorithm can realize high-precision and real-time recognition under the weak light, strong light, uneven illumination, and rotation angle. The success rate is 99.25%, and the time consumed is only 1/4 of the normalized cross-correlation algorithm. The vision positioning and dicing experiment of the semiconductor chips was carried out on a high-precision dicing saw. The results confirm that the improved algorithm could be used for high-precision and real-time dicing semiconductor chips.

CSIR4G: An effective and efficient cross-scenario image retrieval model for glasses

In this work, we propose an effective and efficient cross-scenario glasses retrieval model with the goal of providing a new online shopping experience. Two challenging issues arise for cross-scenario glasses retrieval: identifying glasses in a query image and extracting features from the identified glasses. To properly address these issues, we introduce a novel segmentation-free framework which includes a glasses detection model, an attribute recognition model, and a coarse-to-fine search strategy. Specifically, a new keypoint-based scheme called EyeGlasses keYPoinT (EGYPT) for glasses detection is proposed, in which a set of representative keypoints along a glasses frame are automatically identified. Based on these detected keypoints, various local feature descriptors are extracted to learn the semantic attributes of the glasses. Furthermore, fast retrieval is achieved by applying a hierarchical search strategy based on the index of the attributes. An experimental analysis with real-world photos and product images verifies the efficacy and efficiency of our proposed retrieval model.

Low-rank and sparse decomposition based shape model and probabilistic atlas for automatic pathological organ segmentation.

One major limiting factor that prevents the accurate delineation of human organs has been the presence of severe pathology and pathology affecting organ borders. Overcoming these limitations is exactly what we are concerned in this study. We propose an automatic method for accurate and robust pathological organ segmentation from CT images. The method is grounded in the active shape model (ASM) framework. It leverages techniques from low-rank and sparse decomposition (LRSD) theory to robustly recover a subspace from grossly corrupted data. We first present a population-specific LRSD-based shape prior model, called LRSD-SM, to handle non-Gaussian gross errors caused by weak and misleading appearance cues of large lesions, complex shape variations, and poor adaptation to the finer local details in a unified framework. For the shape model initialization, we introduce a method based on patient-specific LRSD-based probabilistic atlas (PA), called LRSD-PA, to deal with large errors in atlas-to-target registration and low likelihood of the target organ. Furthermore, to make our segmentation framework more efficient and robust against local minima, we develop a hierarchical ASM search strategy. Our method is tested on the SLIVER07 database for liver segmentation competition, and ranks 3rd in all the published state-of-the-art automatic methods. Our method is also evaluated on some pathological organs (pathological liver and right lung) from 95 clinical CT scans and its results are compared with the three closely related methods. The applicability of the proposed method to segmentation of the various pathological organs (including some highly severe cases) is demonstrated with good results on both quantitative and qualitative experimentation; our segmentation algorithm can delineate organ boundaries that reach a level of accuracy comparable with those of human raters.

Hierarchical search strategy in particle filter framework to track infrared target

A target of interest may exhibit significant appearance variations because of its complex maneuvers, ego-motion of the camera platform, etc. Currently, target tracking in forward-looking infrared (FLIR) sequences is still a challenging problem in the field of computer vision. Although many efforts have been devoted, there are still some issues to be addressed. First, state particles generated by prior information cannot approximate the probability density function well when the target state changes obviously. Second, plenty of particles have to be employed to obtain satisfying estimation of target state which will cause heavy computational burden in turn. In this paper, a hierarchical search strategy (HS tracker) is proposed to track infrared target in the particle filter framework, and there are two observation models employed to locate the target robustly. In the first stage, a saliency map leads the redistributed state particles to cover the salient areas that can provide a rough prediction of the target areas. In the second stage, sparse representation is employed to search for a subset of true ones from all the target candidates; thus, only efficient state particles are used to estimate the target state. The proposed method is tested on numerous FLIR sequences from the US army aviation and missile command database, and experimental results demonstrate the excellent performance.

Akbaba—An Agent for the Angry Birds AI Challenge Based on Search and Simulation

In this paper, we report on our entry for the AI Birds competition, where we designed, implemented, and evaluated an agent for the physics puzzle computer game Angry Birds. Our agent uses search and simulation to find appropriate parameters for launching birds. While there are other methods that focus on qualitative reasoning about physical systems we try to combine simulation and adjustable abstractions to efficiently traverse the possibly infinite search space. The agent features a hierarchical search scheme where different levels of abstractions are used. At any level, it uses simulation to rate subspaces that should be further explored in more detail on the next levels. We evaluate single components of our agent and we also compare the overall performance of different versions of our agent. We show that our approach yields a competitive solution on the standard set of levels.

RESEARCH ON THE LAYERING A* ALGORITHM FOR REAL-TIME NAVIGATION

Abstract. According to the existing problems in the applications of embedded navigation, this paper designs the hierarchical search A*algorithm, based on the transferring road network, to meet the need of real-time navigation. In the algorithm, a hierarchical search strategy is applied to route programming of large area, yet the duplicate searching A* algorithm, based on the transferring road network, is applied to the path computation, which is able to handle intersection turn restrictions and node weight, with little storage space but fast searching speed. Practically, the algorithm is proved to meet the technological need of real-time navigation both in computing speed and route rationality.

Paediatric terminology in the Australian health and health-education context: a systematic review.

This study aimed to identify paediatric terminology used in the Australian health and health-education context, propose a standardized framework for Australian use, and compare it with a US-based framework. Australian health and health-education websites were systematically searched using a novel hierarchical domain-specific search strategy to identify grey literature containing paediatric terminology. Webpages published from 2009 to February 2014, with a '.gov.au' or '.edu.au' domain and no advertising, were included. Paediatric terms were analysed with power-law distributions. Age definitions were grouped using a chi-squared test automatic interaction detection analysis (p<0.05). In total, 34 paediatric terms and 197 unique age definitions were identified in 613 webpages. Terms displayed a language distribution, although definitions had semantic and lexical ambiguity. Age definitions were divided into four statistically different groups (F=245.3, p<0.001). Four paediatric terms with distinct age definitions were proposed based on Australian data: 'infant: 0 to <1year', 'early childhood: 1year to <5years', 'child: 5years to <13years', and 'young person: 13years to <22years'. These recommendations were broader than the US-based comparison. This is a starting point for standardizing Australian paediatric terminology, and a method for exploring paediatric terminology in other countries.

Autobiographical memory and hierarchical search strategies in depressed and non-depressed participants.

BackgroundThere is a growing body of literature showing individuals with depression and other trauma-related disorders (e.g., posttraumatic stress disorder) recall more overgeneral and less specific autobiographical memories compared to normal participants. Although the mechanisms underlying overgeneral memory are quite clear, the search strategy operated within the autobiographical knowledge base, at time of recollection, requires further exploration. The current study aimed to examine the hierarchical search sequence used to recall autobiographical memories in depressed and non-depressed participants, with a view to determining whether depressed participants exhibited truncated search strategies.MethodsThirteen depressed and an equal number of non-depressed participants retrieved 15 memories each, in response to 15 commonly used cue words. Participants reported the first memory that entered in their mind. All memory descriptions were recorded and later transcribed verbatim for content analysis.ResultsDepressed participants retrieved autobiographical memories faster, produced shorter memory descriptions and were less likely to recall positive memories than non-depressed participants. Non-depressed participants were more likely to commence retrieval by accessing lifetime period knowledge followed by general event and event specific knowledge, whereas depressed participants showed a tendency to terminate retrieval at the general event level.ConclusionsIt is concluded that depressed participants do adhere to the same hierarchical search strategy as non-depressed participants when retrieving specific autobiographical memories, but that they terminate their search early, resulting in overgeneral memories.

다시점 비디오 부호화를 위한 개선된 예측 구조와 움직임 추정 기법

Multi-view video is obtained by capturing one three-dimensional scene with many cameras at different positions. The computational complexity of multi view video coding increases in proportion to the number of cameras. To reduce computational complexity and maintain the image quality, improved prediction structure and motion estimation method is proposed in this paper. The proposed prediction structure exploits an average distance between the current picture and its reference pictures. The proposed prediction structure divides every GOP into several groups to decide the maximum index of hierarchical B layer and the number of pictures of each B layer. And the proposed motion estimation method uses a hierarchical search strategy. This strategy method consists of modified diamond search pattern, progressive diamond search pattern and modified raster search pattern. Experiment results show that the complexity reduction of the proposed prediction structure and motion estimation method over JMVC (Joint Multiview Video Coding) reference model using hierarchical B pictures of Fraunhofer-HHI and TZ search method can be up to 40~70% while maintaining similar video quality and bit rates.

A Discrete Approach for Pairwise Matching of Archaeological Fragments

This article addresses the problem of automatic reconstruction of ancient artifacts from archaeological fragments. The technique described here focuses on pairwise matching of flat fragments (typically fresco fragments), and it is intended to be the core of a larger system for artifact reconstruction. Global registration techniques are challenging due to the combinatory explosion that happens in the solution space: the goal is to find the best alignment among all possible ones without an initialization. This fact defines the duality between performance and correction that we face in this work. The proposed technique defines a cost function to evaluate the quality of an alignment based on a discrete sampling of the fragments that ensures data alignment. Starting from an exhaustive search strategy, the technique progressively incorporates new features that lead to a hierarchical search strategy. Convergence and correction of the resulting technique are ensured using an optimistic cost function. Internal search calculations are optimized so the only operations performed are additions, subtractions, and comparisons over aligned data. All heavy geometric operations are carried out by the GPU on a preprocessing stage that only happens once per fragment.