ProTSA: A Testing Process for Automotive Software Domain

This study evaluates the development of a testing process for the automotive software domain, highlighting challenges stemming from the absence of adequate processes. The research demonstrates the application of Design Science Research methodology in developing, an automotive software testing process—ProTSA, using six functional testing modules. Additionally, the study evaluates the benefits of implementing ProTSA in a specific Original Equipment Manufacturer (OEM) using an experimental single-case approach with industry professionals’ participation through a survey. The study concludes that combining testing techniques with effective communication and alignment is crucial for enhancing software quality. Furthermore, survey data indicates that implementing ProTSA leads to productivity gains by initiating tests early, resulting in time savings in the testing program and increased productivity for the testing team. Future work will explore implementing ProTSA in cybersecurity, over-the-air software updates, and autonomous vehicle testing processes. .

Development Methodologies for Network Softwarization: A Comparison of DevOps, NetOps, and Verification

This white paper explores three popular development methodologies for network softwarization: DevOps, NetOps, and Verification. The paper compares and contrasts the strengths and weaknesses of each approach and provides recommendations for organizations looking to adopt network softwarization.

A Neuro-Based Software Fault Prediction with Box-Cox Power Transformation

Software fault prediction is one of the most fundamental but significant management techniques in software dependability assessment. In this paper we concern the software fault prediction using a multilayer-perceptron neural network, where the underlying software fault count data are transformed to the Gaussian data, by means of the well-known Box-Cox power transformation. More specially, we investigate the long-term behavior of software fault counts by the neural network, and perform the multi-stage look ahead prediction of the cumulative number of software faults detected in the future software testing. In numerical examples with two actual software fault data sets, we compare our neural network approach with the existing software reliability growth models based on nonhomogeneous Poisson process, in terms of predictive performance with average relative error, and show that the data transformation employed in this paper leads to an improvement in prediction accuracy.

Electrical Metrology Applications of LabVIEW Software

Automation in measurement has wide range of electrical metrology applications and construction of powerful calibration software is one of the highly accurate metrological laboratories’ priorities. Thus, two automatic systems for controlling and calibrating the electrical reference standards have been established at National Institute for Standards (NIS), Egypt. The first system has been built to calibrate the zener diode reference standards while the second one has been built to calibrate the electrical sourcing and measuring instruments. These two systems act as the comprehensive and reliable structure that, from the national electrical standards, disseminates the traceability to all the electrical units under calibration. The software of the two systems has been built using the Laboratory Virtual Instrument Engineering Workbench (LabVIEW) graphical language. The standard development procedures have been followed in the building of both systems software. The software requirement specifications as well as functional specifications are taken into consideration. Design, implementation and testing of the software have been performed. Furthermore, software validation for measurements’ uncertainty as well as results’ compatibility in both automatic and manual modes has been achieved.

Investigating Efficacy of “Working Memory Training Software” on Students Working Memory

The aim of the present study is developing “Working Memory Training Software”, and investigating its content validity and the efficacy of this computerized cognitive training on students working memory. This study is in R & D research category, and it is performed in a semi-experimental design. Its data were collected from students of the third grade (30), fourth grade (30), and fifth grade (12) of primary school. After specifying the software content validity by asking expertise opinions, and investigating these opinions through Spearman Test (rs = 1), these children attended in intervention program for 10 sessions. The subtests of working memory in “Tehran-Stanford-Binet Intelligence Scale” and “Wechsler intelligence scale for children” were conducted in all three groups on the pre-test and post-test. After elimination of the pre-test effect, Paired-Samples T-Test on total scores of subtests of working memory in “Tehran-Stanford Binet Intelligence Scale” (t = 10.869, df = 71, r = 0.967 & P < 0.05) and in “Wechsler intelligence scale for children” (t = 16.809, df = 71, r = 0.983 & P < 0.05) reveals a significant difference in post-test scores. Based on this study the Working Memory Training Software has proper psychometric properties and causes significant improvement in students working memory performance.

Enantioseparation of Palonosetron Hydrochloride and Its Related Enantiomeric Impurities by Computer Simulation and Validation

A rapid, simple and single stereo selective high-performance liquid chromatographic (HPLC) method was developed and validated for enantiomers of palonosetron hydrochloride (PALO) and its process related chiral impurities. A computer simulating software was used for the development of chiral method. The developed method was able to separate not only the enantiomers of palonosetron hydrochloride but also its process related chiral impurities within 12 min. The chromatographic separation was carried out by normal phase chromatography using a 3 μm column of cellulose based chiral stationary phase (Chiralcel-OD 250mm × 4.6mm) with a mobile phase comprised of n-hexane: ethanol: methanol: heptafluoro butyric acid: diethyl amine (70:15:15:0.05:0.1, v/v) at a flow rate of 1.0 mL/min. The effects of additive concentration as well as nature of polar organic modifier, flow rate, and temperature on enantioselectivity were investigated. The limit of detection (LOD) and limit of quantification (LOQ) of the palonosetron isomers and its related chiral impurities were found to be in the range 0.06-0.10 μg/mL and 0.14 - 0.24 μg/mL respectively. The method showed excellent linearity (R2 > 0.998) over a range of 0.14 to 1.125 μg/mL. The percentage recovery of the isomers in bulk drug samples ranged from 87.0 to 116.0.

软件验证与确认技术和标准化研究

软件验证与确认是保证软件质量方面最有效手段之一,对于安全攸关软件的安全性、可靠性的质量提升效果显著,验证与确认两个概念也经常被混淆。从标准化角度对软件验证与确认的概念进行分析和对比,选取典型标准开展分析和阐述,以推动标准在软件生存周期的应用,促进软件质量提升。

城镇排水管网GIS数据校核技术与软件开发

地理信息系统(GIS)是“智慧水务”建设中常见数据库,由于数据库建立、维护等过程的失误,实际的排水管网GIS存在各类异常,这影响了排水管网运行分析、模拟、管理决策的准确性与有效性,故有必要研究排水管网GIS校核技术,以保证GIS数据质量。研究提出了构造拓扑信息算法,总结了拓扑、文本属性、数值属性等常见异常形式,并在此基础上针对各类数据提出通用校核方法,开发了集成排水管网GIS拓扑与属性校核技术以及可输出校核结果报告的软件平台,并将其用于校核Z市污水管网GIS。结果表明,该市排水管网GIS存在24 336条异常记录,经过校核67%以上的异常得到改善,表明这些校核技术可有效地检测与修复各类GIS异常。

工作量感知软件缺陷预测中偏斜分布的影响及测试评估方法

针对工作量感知软件缺陷预测中传统模型测试评估方法存在偏差这一问题,采用偏斜分布的偏度作为数值特征,研究了3种主要测试评估方法的测试集在工作量偏度的偏差和与其对应的估计误差,并基于偏度偏差较小的采样余量方法,提出一种改进方法——后采样方法,所提后采样方法能够保持测试集的类标签比例以避免生成无效测试集。研究结果表明:最常用的十折交叉验证方法偏度偏差最大,其估计误差也最大;与十折交叉验证相比,改进方法性能估计误差减少约4.9%~26.9%;与采样余量方法相比,改进方法不会产生无效测试集,并证明了减小测试集偏度偏差以减少估计误差的有效性。所提后采样方法为工作量感知软件缺陷预测提供了一种更可靠的测试评估方法,能够更准确地评估模型性能。

面向机载软件的预期功能安全分析验证过程及方法研究

预期功能安全(Safety of the Intended Functionality,SOTIF)关注系统与外界环境、交联设备、任务场景和操作人员交互时,由自身功能设计不足而导致的安全隐患,非常适用于具有复杂功能逻辑的系统和软件研制过程。但目前尚未见到SOTIF在机载软件安全性分析验证工作中的研究与应用,导致机载软件安全性分析验证过程难以适用于复杂失效的分析识别。因此借鉴SOTIF在汽车领域的成功应用经验,开展面向机载软件的SOTIF分析验证过程与方法研究。首先,参考ISO 21448标准,提出机载软件SOTIF分析验证框架。然后,借助功能危险分析、故障树模型、场景驱动等理论,针对过程中涉及的SO-TIF分析验证技术进行研究,识别机载系统危险,分析软件异常控制行为及其原因,构建SOTIF测试场景与测试用例,形成基于SOTIF的机载软件安全性分析验证完整闭环。最后,通过SOTIF技术在机轮转弯控制软件的典型工程应用,验证了该研究成果的有效性和可行性,形成了面向机载软件的SOTIF分析验证过程与能力,可支撑研制人员充分识别机载软件运行过程中软硬耦合冲突、人机交互异常、场景切换异常等复杂失效模式,确保机载软件满足高安全、高可靠研制要求。

汽车抬头显示HUD软件验证方法研究

随着汽车智能座舱系统的普及,越来越多的车型会配备汽车抬头显示HUD系统。文章结合某车型抬头显示HUD的开发过程,详述HUD的软件验证方法,并研究软件验证的优化方案。

Iterative Methods for Parametric Linear Systems with Linear Functions

This paper mainly proposes a new C-XSC （C- for eXtended Scientific Computing） software for the symmetric single step method and relaxation method for computing an enclosure for the solution set and compares the methods with others＇ and then makes some modifications and finally, examples illustrating the applicability of the proposed methods are given.

Lessons Learned from Practical Independent Verification and Validation Based on IEEE 1012

IEEE 1012 [1] describes the SDLC phase activities for software independent verification and validation (IV & V) for nuclear power plant in truly general and conceptual manner, which requires the upward and/or downward tailoring on its interpretation for practical IV & V. It contains crucial and encompassing check points and guidelines to analyze the design integrity, without addressing the formalized and the specific criteria for IV & V activities confirming the technical integrity. It is necessary to list up the inspection viewpoint via interpretation of the standard that is practical review points checking design consistency. For fruitful IV & V of Control Element Driving Mechanism Control System (CEDMCS) software for Yonggwang Nuclear Power Plant unit 3 & 4, the specific viewpoints and approach are necessary based on the guidelines of IEEE 1012 to enhance the system quality by considering the level of implementation of the theoretical and the practical IV & V. Additionally IV & V guideline of IEEE 1012 does not specifically provide the concrete measure considering the system characteristics of CEDMCS. This paper provides the seven (7) characteristic criteria for CEDMCS IV & V, and by applying these viewpoints, the design analysis such as function, performance, interface and exception, backward and forward requirement traceability analysis has been conducted. The requirement, design, implementation, and test phase were only considered for IV & V in this project. This article also provides the translation of code to map theoretical verification and validation into practical verification and validation. This paper emphasizes the necessity of the intensive design inspection and walkthrough for requirement phase to resolve the design faults because the IV & V of early phase of SDLC obviously contributes to find out most of critical design inconsistency. Especially for test phase IV & V, it is strongly recommended to prepare the test plan document which is going to be the basis for the test coverage selection and test strategy. This test plan document should be based on the critical characteristics of function and performance of CEDMCS. Also to guarantee the independency of V & V organization participating in this project, and to acquire the full package of design details for IV & V, the systematic approach and efforts with an aspect of management is highlighted among the participants.

核级系统、软件和硬件的V&V研究与应用

软件验证和确认(V&V)技术对于提高核电站安全级仪控系统的软件质量发挥着重要作用。在核安全级数字化控制系统(DCS)中,系统、硬件与软件三者同等重要且密不可分,因此有必要将系统和硬件也纳入V&V的范围。讨论了针对核安全级DCS中系统、软件和硬件的V&V标准体系,提出融合新核安全导则HAD 102/10—2021和新工业标准IEEE Std 1012^(TM)-2016关于生命周期要求的V&V过程模型,并阐述了3类V&V活动之间的交互关系及系统V&V和硬件V&V中的重点任务。该研究成果在自主化核安全级仪控平台和睦系统的研制及某核电站DCS工程典型项目的成功应用中,取得了良好效果。其满足核安全级DCS的质量控制需求,同时也符合HAD 102/10—2021的新监管要求。该研究为我国核行业实施系统和硬件的V&V提供了技术借鉴,有助于提升产品安全性和可靠性,也对其他高可靠性领域系统的质量控制有重要参考意义。

青少年特发性脊柱侧凸有限元模型建立与验证的研究现状及进展

背景:青少年特发性脊柱侧凸在中国发病率高,且病情复杂,对青少年的身心健康具有显著危害。基于患者医学影像建立有限元模型可有效分析脊柱应力,模拟不同治疗方法的恢复效果,为患者治疗方式的选择提供科学依据。但目前有限元模型的建立方式种类繁多,尚未形成统一标准,模型质量参差不齐,验证方法也有待完善。目的:从软件选择、关键结构建模(椎体、椎间盘、韧带和躯干模型)及材料赋值等方面综述近年来青少年特发性脊柱侧凸模型建立的研究现状,从几何形态、已有文献对比、实验验证等角度概述模型验证的研究进展,总结分析有限元模型的不足与发展方向。方法:第一作者于2021年12月应用计算机在Web of Science和中国知网数据库检索2010年1月至今发表的相关文献,以“Adolescent idiopathic scoliosis,Finite Element,Biomechanical”为英文检索词,以“青少年特发性脊柱侧凸、有限元、生物力学”为中文检索词,追溯高引用权威参考文献,最终纳入83篇文献进行综述分析。结果与结论:(1)有限元法能够重建脊柱的三维模型,通过生物力学分析模拟脊柱的生理状态,为青少年特发性脊柱侧凸的保守治疗和手术治疗提供理论依据和方案参考。(2)国内外有限元模型建立方式差别较大,中国学者多采用全脊柱CT检测,通过逆向工程重建椎骨,根据椎骨模型反推椎间盘、韧带、上下终板及关节软骨等结构,国外多使用具有多平面自校准X射线的EOS成像系统建立椎骨模型。(3)尽管模型建立过程中的具体软件选择有所区别,但从软件分类上看均包括医学影像处理软件、三维建模软件和有限元分析软件。(4)关于椎骨、椎间盘的有限元材料赋值已形成较为统一标准,但韧带模型由于单元设置不同,材料赋值也有所差异,而韧带的赋值会显著影响模型的结果。(5)有限元模型的验证方式主要包括与原始医学影像对比的几何验证、与已有文献结果对比的力学验证以及对有限元分析结果的临床试验验证3大类。

基于FPGA的核电厂保护系统V&V方法研究

核电厂保护系统的设计是功能验证和应用软件开发需求的源头,而基于现场可编程门阵列(FPGA)技术开发的保护系统应用则是对保护功能的分解实施和最终实现。针对验证和确认(V&V)全生命周期过程进行了研究,结合保护系统功能设计和基于FPGA技术的功能应用实施,建立了一套完整的V&V方法,明确了每个活动环节的主要任务,提出了面向核电厂保护系统FPGA应用的V&V方法和框架方案。结合目前工程承包项目实施模式,对V&V工作的实施策略给出了优化方向和改进建议。

基于GAN数据增强的软件缺陷预测聚合模型

在软件缺陷预测任务中,通常基于C&K等静态软件特征数据集,使用机器学习分类算法来构建软件缺陷预测(SDP)模型。然而,大多数静态软件特征数据集中缺陷数较少,数据集的类不平衡问题较为严重,导致学习到的SDP模型的预测性能较差。文中基于生成对抗网络(GAN),并利用FID得分筛选生成正例样本数据,增强正例样本量,然后在组块正则化m×2交叉验证(m×2BCV)框架下,通过众数投票法聚合多个子模型的结果,最终构成SDP模型。以PROMISE数据库下的20个数据集为实验数据集,采用随机森林算法构建SDP聚合模型。实验结果表明,与传统的随机上采样、SMOTE、随机下采样相比,所提SDP聚合模型的F1平均值分别提高了10.2%,5.7%,3.4%,且F1的稳定性也得到相应提高;所提SDP聚合模型在20个数据集的评测中,有17个F1值最高。从AUC指标来看,所提方法与传统的采样方法没有明显差异。

GJB 5000B软件测试与监督探析

首先,介绍了GJB 5000B软件测试类型及测试方法,包括单元测试、集成测试、合格性测试、系统测试的测试流程和活动流程;其次,分别对单元测试、集成测试、合格性测试、系统测试的测试目的和测试方法进行了详细介绍;然后,讲解了GJB 5000B软件测试对标的验证和确认过程域的标准要求和关键活动及产品;最后,阐述了实际项目运行中,作为军代表应该如何对软件测试进行日常监督。实践证明,软件测试是软件研制过程的重要环节,可大大地提高软件产品的质量。

核电厂人工智能软件验证和确认方法研究

核电厂仪控领域引入人工智能(AI)技术的同时,也面临着软件验证和确认问题。现有的软件验证和确认(V&V)方法及相关标准并未考虑AI软件的特性。为解决这一问题,研究了AI软件不同于常规软件的特点。提出了在现有V&V活动的基础上,从需求分析、数据集、训练模型、可解释性、底层软件系统以及可维护性这六个方面对AI软件作进一步V&V。同时,给出执行这些任务的时机及需重点考虑的内容。该研究为AI软件在核电厂的应用提供了一种鉴定思路,并拓展了软件V&V的应用范围。