Safety Assessment of Liquid Launch Vehicle Structures Based on Interpretable Belief Rule Base

A liquid launch vehicle is an important carrier in aviation,and its regular operation is essential to maintain space security.In the safety assessment of fluid launch vehicle body structure,it is necessary to ensure that the assessmentmodel can learn self-response rules from various uncertain data and not differently to provide a traceable and interpretable assessment process.Therefore,a belief rule base with interpretability(BRB-i)assessment method of liquid launch vehicle structure safety status combines data and knowledge.Moreover,an innovative whale optimization algorithm with interpretable constraints is proposed.The experiments are carried out based on the liquid launch vehicle safety experiment platform,and the information on the safety status of the liquid launch vehicle is obtained by monitoring the detection indicators under the simulation platform.The MSEs of the proposed model are 3.8000e-03,1.3000e-03,2.1000e-03,and 1.8936e-04 for 25%,45%,65%,and 84%of the training samples,respectively.It can be seen that the proposed model also shows a better ability to handle small sample data.Meanwhile,the belief distribution of the BRB-i model output has a high fitting trend with the belief distribution of the expert knowledge settings,which indicates the interpretability of the BRB-i model.Experimental results show that,compared with other methods,the BRB-i model guarantees the model’s interpretability and the high precision of experimental results.

Deep Belief Network for Lung Nodule Segmentation and Cancer Detection

Cancer disease is a deadliest disease cause more dangerous one.By identifying the disease through Artificial intelligence to getting the mage features directly from patients.This paper presents the lung knob division and disease characterization by proposing an enhancement calculation.Most of the machine learning techniques failed to observe the feature dimensions leads inaccuracy in feature selection and classification.This cause inaccuracy in sensitivity and specificity rate to reduce the identification accuracy.To resolve this problem,to propose a Chicken Sine Cosine Algorithm based Deep Belief Network to identify the disease factor.The general technique of the created approach includes four stages,such as pre-processing,segmentation,highlight extraction,and the order.From the outset,the Computerized Tomography(CT)image of the lung is taken care of to the division.When the division is done,the highlights are extricated through morphological factors for feature observation.By getting the features are analysed and the characterization is done dependent on the Deep Belief Network(DBN)which is prepared by utilizing the proposed Chicken-Sine Cosine Algorithm(CSCA)which distinguish the lung tumour,giving two classes in particular,knob or non-knob.The proposed system produce high performance as well compared to the other system.The presentation assessment of lung knob division and malignant growth grouping dependent on CSCA is figured utilizing three measurements to be specificity,precision,affectability,and the explicitness.

A Processor Performance Prediction Method Based on Interpretable Hierarchical Belief Rule Base and Sensitivity Analysis

The prediction of processor performance has important referencesignificance for future processors. Both the accuracy and rationality of theprediction results are required. The hierarchical belief rule base (HBRB)can initially provide a solution to low prediction accuracy. However, theinterpretability of the model and the traceability of the results still warrantfurther investigation. Therefore, a processor performance prediction methodbased on interpretable hierarchical belief rule base (HBRB-I) and globalsensitivity analysis (GSA) is proposed. The method can yield more reliableprediction results. Evidence reasoning (ER) is firstly used to evaluate thehistorical data of the processor, followed by a performance prediction modelwith interpretability constraints that is constructed based on HBRB-I. Then,the whale optimization algorithm (WOA) is used to optimize the parameters.Furthermore, to test the interpretability of the performance predictionprocess, GSA is used to analyze the relationship between the input and thepredicted output indicators. Finally, based on the UCI database processordataset, the effectiveness and superiority of the method are verified. Accordingto our experiments, our prediction method generates more reliable andaccurate estimations than traditional models.

基于DBN和BES-LSSVM的矿用压风机异常状态识别方法

针对矿用压风机这类分布式系统的异常类别复杂、识别精度低等问题,提出了一种基于深度置信网络(DBN)和最小二乘支持向量机(LSSVM)的异常状态识别方法。首先,分析压风机组成系统及其运行机理,确定常见的异常状态类型;其次,采用DBN无监督学习方式充分挖掘监测数据中异常特征并快速提取;然后,利用秃鹰搜索算法(BES)优化LSSVM的超参数,构建最优的BES-LSSVM分类模型;最后,将DBN提取的异常特征作为BES-LSSVM模型的输入,对矿用压风机异常状态进行识别。试验验证与对比分析结果表明,相较于GA,PSO,GWO算法,BES算法的求解精度和收敛速度均有所提高,同时DBN-BES-LSSVM模型在测试集上平均识别精度达到94.65%,较PCA-LSSVM模型、DBN模型和DBN-LSSVM模型的识别精度分别提高了10.53%,5.84%和3.76%,验证了DBN-BES-LSSVM模型在矿用压风机异常特征提取以及特征识别方面的优越性。

一种基于SSA-DBN的室内可见光指纹定位算法

室内可见光定位在精度方面有着较高的要求,针对这一问题,文中提出了一种麻雀搜索算法(Sparrow Search Algorithm,SSA)优化深度置信网络(Deep Belief Network,DBN)的室内可见光指纹定位算法。首先,采用信号强度特征值与位置坐标建立离线指纹库;其次,利用麻雀搜索算法较好的全局探索和局部开发的能力,对深度置信网络的初始权阈值进行优化,建立网络训练模型,对待定位目标的位置进行预测,避免了DBN陷入局部最优以及收敛速度较慢的问题。最后,利用已建立的离线指纹库数据,计算定位误差并分析。在4 m×4 m×2.5 m的空间中进行实验,结果表明:文中算法的平均定位误差为3.51 cm,定位误差在6 cm以内的概率为89.9%,与DBN定位算法相比,平均定位误差下降了约22.5%。

基于IMODA自适应深度信念网络的复杂模拟电路故障诊断方法

针对传统DBN在无监督训练过程中预训练耗时久、诊断精度差等问题,提出了一种基于改进多目标蜻蜓优化自适应深度信念网络(IMODA-ADBN)的模拟电路故障诊断方法。首先,根据参数更新方向的异同提出了自适应学习率,提高网络收敛速度;其次,传统DBN在有监督调优过程利用BP算法,然而BP算法存在易陷入局部最优的问题,为了改善该问题,利用改进的MODA算法取代BP算法提高网络分类精度。在IMODA算法中,添加Logistic混沌印射和基于对立跳跃以获得帕累托最优解,增加算法的多样性,提高算法的性能。在7个多目标数学基准问题上测试该算法,并与3种元启发式优化算法(MODA、MOPSO和NSGA-II)进行比较,证明了IMODA-ADBN网络模型具有稳定性。最后将IMODAADBN运用到二级四运放双二阶低通滤波器的诊断实验中,实验结果表明该方法在收敛速度快的基础上保证了分类精度,诊断率更高,能够实现高难故障的分类与定位。

基于自适应深度置信网络的压力变送器温度补偿方法研究

随着压力变送器检测技术和人工智能技术的不断发展,在航空航天、石化、核电等领域人们对压力变送器的稳定性、实时性、测量精度等方面有了更严格的要求。而工作环境的温度会对设备精度造成巨大影响,导致变送器测量值出现偏移。针对此问题,本文提出了基于自适应深度置信网络的高精度压力变送器温度补偿方法。深度置信网络(Deep Belief Networks,DBN)在无监督学习阶段提取数据的特征,然后在有监督阶段使用少量的数据对网络参数进行微调;利用白鲸优化算法(Beluga Whale Optimization,BWO)在全局搜索和局部寻优之间达到平衡,有效地提高DBN网络的优化效果;引入Metropolis准则和适应度平衡因子,进一步提高算法的全局寻优能力以及模型收敛速度。实验拟合后的数据精度可达0.0048%,远高于现有的最高标准0.05级。经过一系列对比分析,验证了补偿算法的准确性和实用性。

3×3核矩阵极化码的BP译码算法

相比于2×2核极化码,3×3核极化码的码长更加丰富以及有着更高的极化速率。同时,极化码的置信传播(Belief Propagation, BP)算法相比于传统串行消去(Successive Cancellation, SC)译码算法具有更低的延时。将2×2核极化码的BP译码算法推广至3×3核。获得了3×3核内部最小计算单元的信息更新公式;基于信息更新公式,给出了3×3核的BP译码算法。仿真结果表明,对于3×3核极化码,BP译码算法相比于SC译码算法,在中低信噪比下性能要优于SC译码算法。在译码性能相当的条件下,BP译码算法有着更低的译码延时。

Steiner树优化问题的算法研究综述

最优Steiner树问题(Steiner tree problem,STP)是一个经典的组合优化问题,许多工程问题都可以归结为最优Steiner树问题。STP被广泛应用于通信网络、电路设计、VLSI设计等领域。然而,STP是典型的NP难问题,还没有多项式时间的精确算法求解该问题。目前,求解该问题的算法主要集中在基于启发式的近似算法、智能优化算法、信息传播算法等,并取得了很好的效果。在不同规模的网络中,基于传统遗传算法给出一种叶交叉机制(leaf crossover,LC),使用该机制的算法性能表现更好。通过对这些算法的原理、性能、精度等方面进行梳理,归纳出算法的优缺点,并指出STP的研究方向和算法设计路径,对于相关问题的研究有指导意义。

基于CS-DBN的锂电池剩余寿命预测

为了更准确地对锂电池剩余使用寿命进行预测,提出一种基于布谷鸟算法(CS)和深度信念网络(DBN)的预测模型。首先,引进16个影响锂电池RUL的健康因子(HI),通过随机森林(RF)选择出对于剩余寿命预测较为重要的9个HI。随后用CS去寻优深度信念网络模型中隐藏层的参数,通过寻优,建立最优的深度信念网络预测模型。最后,使用马里兰大学所收集的电池数据(CALCE)进行实验,结果表明:所提出的CS-DBN模型的拟合优度高达98%,且与其他模型的预测结果进行对比,具有更小的误差,验证了所提方法的有效性。

分组衰落信道下LDPC的一种带信道估计的改进Belief-Propagation算法

该文首先通过仿真证明了LDPC(Low Density Parity Check code)在分组衰落信道下,以通常的 Belief-propagation算法译码,具有较好性能。然后基于算法的特殊迭代特性,提出分组衰落信道下,在每一迭代步 骤中结合信道估计的改进的Belief-propagation算法。仿真证明,该算法可以有效地减少译码迭代次数。

基于GA-IPSO-KPCA和变权组合模型的电动汽车充电方法

需求电压和需求电流是充电桩对电动汽车安全充电的重要依据。然而,随着电池的老化,电池管理系统的数据可能出现错误,使得电动汽车在充电时存在安全隐患。针对该问题,建立最小二乘支持向量机和深度置信网络的组合预测模型,提出一种基于变权组合模型的电动汽车充电方法。首先,针对数据掉线缺失问题,使用K均值和反距离加权方法对数据进行插值;然后,使用改进的混合核主成分分析算法对完整数据进行主成分提取,并使用改进粒子群优化算法自动确定混合核函数的权重。基于真实电动汽车数据的实验结果表明,所提方法能够准确地预测需求电压和需求电流,具有实际意义和可行性。

基于WPT-ARO-DBN/WPT-EPO-DBN模型的月含沙量多步预测

准确的含沙量多步预测对于区域水土流失治理、防洪减灾等具有重要意义。为提高含沙量多步预测精度,改进深度信念网络(DBN)的预测性能,基于小波包变换(WPT),分别提出人工兔优化(ARO)算法、鹰栖息优化(EPO)算法与DBN组合的月含沙量多步预测模型,通过云南省龙潭站月含沙量时序数据对模型进行验证。首先利用WPT对实例月含沙量时序数据进行3层分解处理,得到8个更具规律的子序列分量;其次介绍ARO、EPO算法原理,利用ARO、EPO对DBN隐藏层神经元数等超参数进行寻优,建立WPT-ARO-DBN、WPT-EPO-DBN预测模型,并构建WPT-PSO(粒子群算法)-DBN、WPT-DBN作对比分析模型;最后利用4种模型对各子序列分量进行预测,将预测值叠加得到最终月含沙量多步预测结果。结果表明:(1)WPT-ARO-DBN、WPT-EPO-DBN模型对实例超前1步—超前4步月含沙量具有满意的预测效果,对超前5步具有较好的预测结果,对超前6步、超前7步的预测效果一般,对超前8步的预测精度较差,已不能满足预测精度需求;(2)WPT-ARO-DBN、WPT-EPO-DBN模型的多步预测效果要优于WPT-PSO-DBN模型,远优于WPT-DBN模型,具有更高的预测精度、更好的泛化能力和更大的预测步长;(3)ARO、EPO能有效优化DBN超参数,提高DBN预测性能,优化效果优于PSO,WPT-ARO-DBN、WPT-EPO-DBN模型能充分发挥WPT、新型群体智能算法和DBN网络优势,提高月含沙量多步预测精度,且预测精度随着预测步数的增加而降低。

铁路信号机房外部冲击电流监测原理及软件分析策略

针对铁路信号机房应用场景下常见的外部冲击电流问题,从基本监测原理、软件分析策略两方面提出具体解决方案,基于该方案可实现对多频段冲击电流的综合采集以及冲击电流详情自动化分析,实现对设备管理单位的指导性建议,提升铁路信号机房运行的安全性。

Optimizing Deep Learning Parameters Using Genetic Algorithm for Object Recognition and Robot Grasping

The performance of deep learning(DL)networks has been increased by elaborating the network structures. However, the DL netowrks have many parameters, which have a lot of influence on the performance of the network. We propose a genetic algorithm(GA) based deep belief neural network(DBNN) method for robot object recognition and grasping purpose. This method optimizes the parameters of the DBNN method, such as the number of hidden units, the number of epochs, and the learning rates, which would reduce the error rate and the network training time of object recognition. After recognizing objects, the robot performs the pick-andplace operations. We build a database of six objects for experimental purpose. Experimental results demonstrate that our method outperforms on the optimized robot object recognition and grasping tasks.

Improved Reduced Latency Soft-Cancellation Algorithm for Polar Decoding

Soft-cancellation(SCAN) is a soft output iterative algorithm widely used in polar decoding. This algorithm has better decoding performance than reduced latency soft-cancellation(RLSC) algorithm, which can effectively reduce the decoding delay of SCAN algorithm by 50% but has obvious performance loss. A modified reduced latency soft-cancellation(MRLSC) algorithm is presented in the paper. Compared with RLSC algorithm, LLR information storage required in MRLSC algorithm can be reduced by about 50%, and better decoding performance can be achieved with only a small increase in decoding delay. The simulation results show that MRLSC algorithm can achieve a maximum block error rate(BLER) performance gain of about 0.4 dB compared with RLSC algorithm when code length is 2048. At the same time, compared with the performance of several other algorithms under(1024, 512) polar codes, the results show that the throughput of proposed MRLSC algorithm has the advantage at the low and medium signal-to-noise ratio(SNR) and better BLER performance at the high SNR.

Convergence Rate Analysis of Gaussian Belief Propagation for Markov Networks

Gaussian belief propagation algorithm(GaBP) is one of the most important distributed algorithms in signal processing and statistical learning involving Markov networks. It is well known that the algorithm correctly computes marginal density functions from a high dimensional joint density function over a Markov network in a finite number of iterations when the underlying Gaussian graph is acyclic. It is also known more recently that the algorithm produces correct marginal means asymptotically for cyclic Gaussian graphs under the condition of walk summability(or generalised diagonal dominance). This paper extends this convergence result further by showing that the convergence is exponential under the generalised diagonal dominance condition,and provides a simple bound for the convergence rate. Our results are derived by combining the known walk summability approach for asymptotic convergence analysis with the control systems approach for stability analysis.

A Software Risk Analysis Model Using Bayesian Belief Network

The uncertainty during the period of software project development often brings huge risks to contractors and clients. If we can find an effective method to predict the cost and quality of software projects based on facts like the project character and two-side cooperating capability at the beginning of the project,we can reduce the risk. Bayesian Belief Network(BBN) is a good tool for analyzing uncertain consequences, but it is difficult to produce precise network structure and conditional probability table.In this paper,we built up network structure by Delphi method for conditional probability table learning,and learn update probability table and nodes’confidence levels continuously according to the application cases, which made the evaluation network have learning abilities, and evaluate the software development risk of organization more accurately.This paper also introduces EM algorithm, which will enhance the ability to produce hidden nodes caused by variant software projects.

Jointly-check iterative decoding algorithm for quantum sparse graph codes

For quantum sparse graph codes with stabilizer formalism, the unavoidable girth-four cycles in their Tanner graphs greatly degrade the iterative decoding performance with standard belief-propagation （BP） algorithm. In this paper, we present a jointly-check iterative algorithm suitable for decoding quantum sparse graph codes efficiently. Numerical simulations show that this modified method outperforms standard BP algorithm with an obvious performance improvement.

An EEGA-Based Bayesian Belief Network Model for Recognition of Human Activity in Smart Home

With the emerging of sensor networks, research on sensor-based activity recognition has attracted much attention. Many existing methods cannot well deal with the cases that contain hundreds of sensors and their recognition accuracy is requisite to be further improved. A novel framework for recognizing human activities in smart home was presented. First, small, easy-to-install, and low-cost state change sensors were adopted for recording state change or use of the objects. Then the Bayesian belief network (BBN) was applied to conducting activity recognition by modeling statistical dependencies between sensor data and human activity. An edge-encode genetic algorithm (EEGA) approach was proposed to resolve the difficulties in structure learning of the BBN model under a high dimension space and large data set. Finally, some experiments were made using one publicly available dataset. The experimental results show that the EEGA algorithm is effective and efficient in learning the BBN structure and outperforms the conventional approaches. By conducting human activity recognition based on the testing samples, the BBN is effective to conduct human activity recognition and outperforms the naive Bayesian network (NBN) and multiclass naive Bayes classifier (MNBC).