Poor Convergence Research Articles

Development of single measuring point overall balancing method based on multi-cylinder dynamic balance detection system

The combine harvester with multi-cylinder has a wide application in improving the operation efficiency. Therefore, as an important process in combine harvester, threshing operated by multi-cylinder usually makes the working structures suffer hit from the blockage of stones or materials. Consequently, the cylinders fail to guarantee the original balance. Several studies have been performed in this area, based on the dynamic balance method and vibration signal processing. However, most of them are not efficient for multi-cylinder. This study aims at developing an online dynamic balance detection system for multi-cylinder in combine harvester, in order to enhance the efficiency and accuracy of online dynamic balance detection. A multi-cylinder test rig is developed, and unbalanced vibration experiments under different working conditions are performed. The results show that the different measuring positions have different sensitivities to unbalanced mass. The vibrations will be transmitted through the frame, and coupled with each other. Under the multi-cylinder rotation, the vibration has strong randomness and poor convergence. For this purpose, the single measuring point overall balancing method is proposed. An adaptive multiple filter algorithm relying on the LabVIEW platform, is also proposed. Through phase relation between cylinders, the single measuring point overall balancing algorithm is developed. The comparison experimental results show that the proposed algorithm is more robust under the multi-cylinder rotation. The proposed method simultaneously installs sensors on the bearing seat of No. 1 tangential cylinder and balance 3 cylinders. The traditional method balance 3 cylinders separately change the sensor position. The effect of the proposed method is almost 76%, while that of the traditional balance method is almost 66%. For the multi-cylinder system, the single measuring point overall balancing method can efficiently improve the balance effect by 10%. Finally, it can be deduced from the experiment on the combine harvester that the accuracy of the proposed method reaches G3.9.

Improving sign-algorithm convergence rate using natural gradient for lossless audio compression

In lossless audio compression, the predictive residuals must remain sparse when entropy coding is applied. The sign algorithm (SA) is a conventional method for minimizing the magnitudes of residuals; however, this approach yields poor convergence performance compared with the least mean square algorithm. To overcome this convergence performance degradation, we propose novel adaptive algorithms based on a natural gradient: the natural-gradient sign algorithm (NGSA) and normalized NGSA. We also propose an efficient natural-gradient update method based on the AR(p) model, which requires mathcal {O}(p) multiply–add operations at every adaptation step. In experiments conducted using toy and real music data, the proposed algorithms achieve superior convergence performance to the SA. Furthermore, we propose a novel lossless audio codec based on the NGSA, called the natural-gradient autoregressive unlossy audio compressor (NARU), which is open-source and implemented in C. In a comparative experiment with existing, well-known codecs, NARU exhibits superior compression performance. These results suggest that the proposed methods are appropriate for practical applications.

Fault data expansion method of permanent magnet synchronous motor based on Wasserstein-generative adversarial network

Aiming at the characteristics of non-smooth, non-linear, multi-source heterogeneity, low density of value and unevenness of fault data collected by the online monitoring equipment of permanent magnet synchronous motor (PMSM), and the difficulty of fault mechanism analysis, this paper proposes a method of PMSM data expansion based on the improved generative adversarial network. First, use the real fault data of the motor to train the model to obtain a mature and stable generative countermeasure network. Secondly, use the generative countermeasure network model to test the remaining data and generate pseudo samples. Finally, use the two-dimensional data analysis method and the time-domain analysis method to generate validity analysis of samples. Aiming at the characteristics of unbalanced motor data, the data expansion method of inter-turn short-circuit faults is carried out based on the data expansion method of the improved generative countermeasure network, and the two-dimensional data analysis method and the time-domain analysis method are used for analysis. The experimental results show that the improved Wasserstein-Generative Adversarial Network (W-GAN) has a better ability to generate fake data, which provides a data basis for the mechanism analysis and machine fault diagnosis of PMSMs. Data analysis results show that the improved W-GAN effectively solves the problem of poor convergence of GAN.

A reinforced exploration mechanism whale optimization algorithm for continuous optimization problems

Whale optimization algorithm (WOA) is a meta-heuristic algorithm. In process of optimizing single-peak problem and multi-peak problem, it has the problems of weak exploration ability, poor convergence behavior and easy to fall into local optimum (LO). To solve these problems, we analyze WOA from the two aspects of global exploration efficiency and convergence characteristics. Correspondingly, the population redistribution and convergent adaptive weighting strategy are proposed. The population redistribution strategy can improve the space utilization of WOA in search process, enhance its exploration efficiency and prevent it from falling into LO. The purpose of the convergent adaptive weighting strategy is to improve convergence behavior and global exploitation efficiency. Experiment result shows that it can effectively enhance the exploration efficiency of WOA. The paper introduces the two strategies to reconstruct a reinforced exploration mechanism whale optimization algorithm (REM-WOA). In order to verify its performance, 36 well known benchmark functions are selected for experiment from the CEC2017. Compared with other 12 meta-heuristic algorithms, the REM-WOA shows obvious advantages. Compared with 9 WOA and its variant algorithms, the REM-WOA has the best convergence behavior and strong global exploration efficiency. For three real design cases studies, REM-WOA has the best exploration efficiency.

Agnostic multimodal brain anomalies detection using a novel single-structured framework for better patient diagnosis and therapeutic planning in clinical oncology

The application of image processing in medical image analysis offers physicians numerous advantages in diagnosing and predicting patient recovery. A typical task that requires sensitive handling is the detection of a cerebral aneurysm, which generally involves the arteries in the human brain. Due to peculiar and complicated tissue structures, such as white and grey matter, which make up most of the brain, it is indeed challenging to visualize the aneurysm and the affected parts. Magnetic resonance angiography (MRA) is a tool for such a process, and it is difficult for radiologists and oncologists to decide. To overcome these drawbacks, the researchers proposed a novel algorithm named artificial bee colony optimization (ABC) with spatially constrained adaptively regularized kernel function-based fuzzy C-means (ABC-ScARKFCM) in this work. The system outperforms the conventional fuzzy C-means clustering method (FCM), which has inaccuracies in intensity handling and segmentation, and a poor convergence rate. The developed algorithm performed well on clinical MRA and Magnetic Resonance images (MRI) from the BraTS challenge dataset (2013, 2015, 2018, 2019, 2020 and 2021). The algorithm achieved dice score, sensitivity and specificity of 87.89%, 98.9% and 98.98%, respectively, which is very remarkable and shows that the applicability of the algorithm can be extended to oncology applications, where suppression of openness/anonymity is expected in diagnosis and assessment of prognosis of patients after therapy.

An adaptive switching-based evolutionary algorithm for many-objective optimization

Pareto-based evolutionary algorithms are challenging in dealing with many-objective problems encountering many incomparable nondominated solutions. To reduce selection pressure and improve diversity, this paper proposes an adaptive switching strategy-based evolutionary algorithm for many-objective optimization. This strategy contains two deletion criteria, which are switched adaptively between generations, aiming to delete poor solutions one by one in environmental selection. The first criterion is devised to delete the solution with poor convergence among the two most similar solutions. The second criterion is developed to delete the worse solution according to a designed indicator that takes into account both convergence and diversity. Finally, comparisons with five state-of-the-art many-objective evolutionary algorithms on some widely used benchmark problems and the water resource planning problem are given to illustrate the effectiveness and advantages of the proposed algorithm.

Energy management strategy via maximum entropy reinforcement learning for an extended range logistics vehicle

The modern energy management strategy (EMS) plays a vital role in the energy efficiency of the extended range electric vehicle. However, some modern strategies such as model predictive control (MPC) and dynamic programming (DP) have limited practical potential because they are subject to the pre-known environment information and noise interference. The reinforcement learning (RL)control strategy can be adopted as online control to interact with the vehicle and the environment. In this study, a novel auxiliary power unit (APU) charging strategy with multi-object optimization is proposed to achieve high fuel conversion efficiency while maintaining battery charging health. The state-of-the-art algorithm, Soft Actor-Critic (SAC), is applied to achieve better exploration of the possible APU behaviour and solve the sensitivity and poor convergence problems from the current RL studies. Its performance is further verified by the results of the Deep Deterministic Policy Gradient (DDPG) algorithm and DP. Three innovative targets are selected as the RL rewards for optimization: the engine fuel rate, SOC charging trajectory, and the battery charging rate (C-rate). The first adoption of the battery C-rate monitoring in RL-based energy management strategy helps extend the battery lifespan from excessive discharge. The comparative results show that the SAC had a 36% faster convergence speed than DDPG while providing a smoother and more stable action space. The fuel consumption with SAC also outplays that of DDPG by around 3%, which achieves almost 95% of the global optimization result. The successful deployment of the SAC algorithm as an EMS indicates its standout ability in dealing with wide-range actions and states with high randomness, revealing the practical potential compared with the existing RL strategies.

An Improved Hill Climbing Algorithm for Graph Partitioning

Abstract Graph partitioning is an NP-hard combinatorial optimization problem, and is a fundamental step in distributing workloads on parallel compute systems, circuit placement, and sparse matrix reordering. The proposed heuristic algorithms such as streaming graph partitioning provide solutions to large-scale graph in a reasonable amount of time. However, the ability of breaking out of local minima in existing these methods is very limited as they are simple in reflecting the connectivity between vertices in real graphs with power-law distribution characteristic. As hill climbing algorithm is a local search method, it can be adopted to improve the result of graph partitioning. However, directly adopting the existing hill climbing algorithm to graph partitioning will result in local minima and poor convergence speed during the iterative process. In this paper, we propose an improved hill climbing graph partitioning algorithm based on clustering. Instead of taking a single vertex as a basic unit, the proposed method considers a cluster consisting of a series of vertices as a hill to move during each iteration. The method uses a new metric that considers both balance and edgecuts to look for the most beneficial cluster as the hill. With these improvements, the method provides a strong power to break out of local minima and achieve an adaptive tradeoff between balance and edgecuts. Experimental results on real-world graphs show that the proposed algorithm substantially reduces edgecuts within a controlled imbalance range.

Optimizing the Calculation of Free Energy Differences in Nonequilibrium Work SQM/MM Switching Simulations.

A key step during indirect alchemical free energy simulations using quantum mechanical/molecular mechanical (QM/MM) hybrid potential energy functions is the calculation of the free energy difference ΔAlow→high between the low level (e.g., pure MM) and the high level of theory (QM/MM). A reliable approach uses nonequilibrium work (NEW) switching simulations in combination with Jarzynski’s equation; however, it is computationally expensive. In this study, we investigate whether it is more efficient to use more shorter switches or fewer but longer switches. We compare results obtained with various protocols to reference free energy differences calculated with Crooks’ equation. The central finding is that fewer longer switches give better converged results. As few as 200 sufficiently long switches lead to ΔAlow→high values in good agreement with the reference results. This optimized protocol reduces the computational cost by a factor of 40 compared to earlier work. We also describe two tools/ways of analyzing the raw data to detect sources of poor convergence. Specifically, we find it helpful to analyze the raw data (work values from the NEW switching simulations) in a quasi-time series-like manner. Principal component analysis helps to detect cases where one or more conformational degrees of freedom are different at the low and high level of theory.

Giant density-of-states van Hove singularities in the face-centered cubic lattice

All van Hove singularities in the density of states (DOS) of face-centered cubic lattice in the nearest and next-nearest neighbor approximation, focusing on higher-order ones, are found and classified. At special values of the ratio τ of nearest (t) and next-nearest neighbor hopping integral t′ giant DOS singularities, caused by van Hove lines or surfaces, are formed. An exact formula for DOS which provides efficient numerical implementation is proposed. The standard tetrahedron method is demonstrated to be inapplicable due to its poor convergence in the vicinity of kinks caused by van Hove singularities. A comparison with the case of large space dimensionality (infinite coordination number) including next-nearest neighbors is performed.

High-Efficiency Optimization Algorithm of PMEPR for OFDM Integrated Radar and Communication Waveform Based on Conjugate Gradient

The tone reservation (TR) approach is mainly adopted to reduce the peak-to-mean envelope power ratio (PMEPR) of orthogonal frequency division multiplexing (OFDM) waveform with low TR ratio (TRR) in classic 4G communication systems. However, for the OFDM integrated radar and communication waveform, high TRR is necessary to simultaneously maintain the radar detection performance as well as communication performance. For cases with high TRR, the traditional waveform optimization algorithms have low execution efficiency and a poor PMEPR convergence level, and thus a new algorithm is needed. This paper proposes a new PMEPR optimization algorithm based on conjugate gradient. Firstly, by introducing the concept of Lp-norm, the PMEPR of the OFDM waveform is accurately remodeled as the objective function of the waveform optimization problem. Secondly, the conjugate gradient of the objective function is analytically derived to form the basis of the efficient PMEPR optimization. Finally, a PMEPR optimization algorithm based on the Polak–Ribière–Polyak (PRP) conjugate gradient is proposed. The simulation results verified the proposed algorithm in terms of optimization efficiency, as well as convergence level, and the initial experimental results suggested the practicality of the proposed algorithm.

Multi-objective microgrid optimal dispatching based on improved bird swarm algorithm

Multi-objective optimal dispatching schemes with intelligent algorithms are recognized as effective measures to promote the economics and environmental friendliness of microgrid applications. However, the low accuracy and poor convergence of these algorithms have been challenging for system operators. The bird swarm algorithm (BSA), a new bio- heuristic cluster intelligent algorithm, can potentially address these challenges; however, its computational iterative process may fall into a local optimum and result in premature convergence when optimizing small portions of multi-extremum functions. To analyze the impact of a multi-objective economic–environmental dispatching of a microgrid and overcome the aforementioned problems of the BSA, a self-adaptive levy flight strategy-based BSA (LF–BSA) was proposed. It can solve the dispatching problems of microgrid and enhance its dispatching convergence accuracy, stability, and speed, thereby improving its optimization performance. Six typical test functions were used to compare the LF–BSA with three commonly accepted algorithms to verify its excellence. Finally, a typical summer-time daily microgrid scenario under grid-connected operational conditions was simulated. The results proved the feasibility of the proposed LF–BSA, effectiveness of the multi- objective optimization, and necessity of using renewable energy and energy storage in microgrid dispatching optimization.

Physical Exercise Monitoring System Based on Computer Somatosensory Technology

Aiming at the problems of poor convergence, slow acceleration, and low throughput in traditional physical exercise monitoring systems, a physical exercise monitoring system based on computer somatosensory technology is proposed and designed. The whole structure of the physical activity monitoring system was created from the hardware and software, according to the research. Human body posture monitoring sensor module, communication connection module, power supply module, main control unit module, sensor module, and other pieces make up the hardware section. To improve detection performance, computer somatosensory technology is used in the software design to gather physical activity posture data. Also, for experimental comparison and verification, we employ system convergence, acceleration, and throughput as indicators. Experimental results show that the improved method can effectively improve system convergence, increase system acceleration, and increase system throughput, which has certain advantages in use.

A boosted chimp optimizer for numerical and engineering design optimization challenges.

Chimp optimization algorithm (ChoA) has a wholesome attitude roused by chimp’s amazing thinking and hunting ability with a sensual movement for finding the optimal solution in the global search space. Classical Chimps optimizer algorithm has poor convergence and has problem to stuck into local minima for high-dimensional problems. This research focuses on the improved variants of the chimp optimizer algorithm and named as Boosted chimp optimizer algorithms. In one of the proposed variants, the existing chimp optimizer algorithm has been combined with SHO algorithm to improve the exploration phase of the existing chimp optimizer and named as IChoA-SHO and other variant is proposed to improve the exploitation search capability of the existing ChoA. The testing and validation of the proposed optimizer has been done for various standard benchmarks and Non-convex, Non-linear, and typical engineering design problems. The proposed variants have been evaluated for seven standard uni-modal benchmark functions, six standard multi-modal benchmark functions, ten standard fixed-dimension benchmark functions, and 11 types of multidisciplinary engineering design problems. The outcomes of this method have been compared with other existing optimization methods considering convergence speed as well as for searching local and global optimal solutions. The testing results show the better performance of the proposed methods excel than the other existing optimization methods.

Modified Remora Optimization Algorithm for Global Optimization and Multilevel Thresholding Image Segmentation

Image segmentation is a key stage in image processing because it simplifies the representation of the image and facilitates subsequent analysis. The multi-level thresholding image segmentation technique is considered one of the most popular methods because it is efficient and straightforward. Many relative works use meta-heuristic algorithms (MAs) to determine threshold values, but they have issues such as poor convergence accuracy and stagnation into local optimal solutions. Therefore, to alleviate these shortcomings, in this paper, we present a modified remora optimization algorithm (MROA) for global optimization and image segmentation tasks. We used Brownian motion to promote the exploration ability of ROA and provide a greater opportunity to find the optimal solution. Second, lens opposition-based learning is introduced to enhance the ability of search agents to jump out of the local optimal solution. To substantiate the performance of MROA, we first used 23 benchmark functions to evaluate the performance. We compared it with seven well-known algorithms regarding optimization accuracy, convergence speed, and significant difference. Subsequently, we tested the segmentation quality of MORA on eight grayscale images with cross-entropy as the objective function. The experimental metrics include peak signal-to-noise ratio (PSNR), structure similarity (SSIM), and feature similarity (FSIM). A series of experimental results have proved that the MROA has significant advantages among the compared algorithms. Consequently, the proposed MROA is a promising method for global optimization problems and image segmentation.

Decision-Making Model of Social Work Specialization Process Using Big Data Analysis and Chaos Cloud

Big data and mobile edge computing information technology have revolutionized traditional industries’ development. The organic integration of community social work and big data technology will undoubtedly promote the growth of community social work services, thanks to the new information technology platform for big data. In terms of the future of big data-driven social work, it is primarily reflected in the following: in the application of social work, big data can acquire and update real human behavior data in real time, and eliminating the need for subjects’ subjective reports in the data collection process; It can also provide a basis for public decision-making, which is worthy of our active exploration and innovation, provide efficient and accurate technical means for risk population assessment and crisis intervention and have a more accurate and intuitive grasp of the temporal and spatial distribution of service personnel and basic social service facilities. The core of professionalization of social work is improving the post-setting system for social workers in our country, improving their treatment, and strengthening the construction of professional ethics for social workers in our country. A big data social workflow adaptive scheduling optimization algorithm based on the chaotic cloud diversion mechanism is proposed to address the problems of high costs, poor time convergence performance, and poor cost accounting quality in the current social workflow adaptive scheduling optimization algorithm. As an important part of social governance, social work accelerates the process of social work specialization to improve decision-making efficiency.

Connection element method for reservoir numerical simulation

<p indent=0mm>Abstract: The reservoir numerical simulation presently encounters problems based on the fine grid. These include poor convergence and difficult history matching. In practical application, the coarse grid often leads to the destruction of the flow structure and the significant reduction of the prediction accuracy. This paper proposes a new reservoir numerical simulation method different from the traditional grid-based methods. The proposed method discretizes the reservoir calculation domain into a series of connection elements from the flow perspective. According to the physical properties of nodes, the support domain and the moving least squares are used to estimate the Laplace operator of pressure and provide the calculation method of the attribute parameters reflecting the flow ability (connection transmissibility) and mass basis (connection volume) of the connection element. Taking the connection element as the object, the flow flux in the element is obtained by solving the node pressure equation. The transport equation about saturation is also solved by using a semianalytical method along the connection element to obtain the production dynamics. The path tracking algorithm is introduced to quantitatively represent the flow interaction and connectivity between the source and sink nodes. Finally, a reservoir numerical simulation method based on the connection element system, which is referred to herein as the connection element method (CEM), is introduced. Practical applications show that the CEM can realize the fast prediction and history matching of the reservoir production performance and identification of the injection-production connectivity. The conceptual example also verifies that in the case of the same node (grid) distribution, the CEM shows better calculation accuracy and efficiency compared with the traditional finite difference method. At the same time, the node allocation in the CEM is more flexible and can ensure the flow path integrity under fewer nodes to improve the prediction accuracy. It also supports the parallel computing of transport equations on connection elements. In conclusion, the CEM can better meet the requirements of the rapid prediction of actual large-scale reservoirs and provide a new idea for reservoir numerical simulation with complex geometric characteristics, such as fracture, fracture-cavity type, and complex boundary.

Research on Economic Optimal Dispatching of Microgrid Cluster Based on Improved Butterfly Optimization Algorithm

Microgrid optimal dispatching has become one of the core issues of microgrid energy management and integrated control, which is of great significance to reduce energy consumption and environmental pollution. As a natural heuristic algorithm, the butterfly optimization algorithm (BOA) has the advantages of simple adjustment parameters and fast convergence speed. It is widely used to solve nonlinear programming problems. However, BOA is easy to fall into local optimization and poor convergence accuracy. Therefore, an improved butterfly optimization algorithm (IBOA) based on skew tent chaotic map, Cauchy mutation, and simplex method is proposed, and compared with particle swarm optimization (PSO), whale optimization algorithm (WOA), sparrow search algorithm (SSA), and BOA, the results show that the IBOA has high convergence speed and optimization accuracy. Finally, the IBOA is used to solve the optimization model. The simulation results show that the IBOA can effectively reduce the power consumption cost of the system, promote the effective utilization of renewable energy, and improve the operation stability of the microgrid cluster system.

Constructing periodic orbits of high-dimensional chaotic systems by an adjoint-based variational method.

Chaotic dynamics in systems ranging from low-dimensional nonlinear differential equations to high-dimensional spatiotemporal systems including fluid turbulence is supported by nonchaotic, exactly recurring time-periodic solutions of the governing equations. These unstable periodic orbits capture key features of the turbulent dynamics and sufficiently large sets of orbits promise a framework to predict the statistics of the chaotic flow. Computing periodic orbits for high-dimensional spatiotemporally chaotic systems remains challenging as known methods either show poor convergence properties because they are based on time-marching of a chaotic system causing exponential error amplification, or they require constructing Jacobian matrices which is prohibitively expensive. We propose a new matrix-free method that is unaffected by exponential error amplification, is globally convergent, and can be applied to high-dimensional systems. The adjoint-based variational method constructs an initial value problem in the space of closed loops such that periodic orbits are attracting fixed points for the loop dynamics. We introduce the method for general autonomous systems. An implementation for the one-dimensional Kuramoto-Sivashinsky equation demonstrates the robust convergence of periodic orbits underlying spatiotemporal chaos. Convergence does not require accurate initial guesses and is independent of the period of the respective orbit.

Deep active contours using locally controlled distance vector flow

Active contours model (ACM) has been extensively used in computer vision and image processing. In recent studies, convolutional neural networks (CNNs) have been combined with ACM replacing the user in the process of contour evolution and image segmentation to eliminate limitations associated with ACM dependence on energy functional parameters and initialization. However, prior studies did not aim for automatic initialization, which is addressed in this article. In addition to manual initialization, current methods are highly sensitive to the initial location and fail to delineate borders accurately. We propose a fully automatic image segmentation method to address problems of manual initialization, insufficient capture range, and poor convergence to boundaries, in addition to the problem of assignment of energy functional parameters. We train two CNNs, one of which generating ACM weighting parameters and the other generating a ground truth mask to extract distance transform (DT) and an initialization circle. DT is used to form a vector field pointing from each pixel of the image towards the closest ground truth boundary point. Vector magnitudes are equal to the Euclidean distance between each pixel and the closest ground truth boundary point. We evaluate our method on four publicly available datasets, including two building instance segmentation datasets, i.e., Vaihingen and Bing huts, and two mammography image datasets, INBreast and DDSM-BCRP. Our approach achieves state-of-the-art results in mean Intersection over Union (mIoU), Dice similarity coefficient and Boundary F-score (BoundF) with the values of 92.33%, 92.44%, and 86.57% for Vaihingen dataset, and 87.12%, 86.86%, and 66.91% for Bing huts dataset. We obtained the Dice similarity coefficient values of 94.23% and 90.89% for the INBreast and DDSM-BCRP, respectively.