Real Machine Research Articles

Comparative tribocorrosion tests of 30CrMo12 cast steel and ADI spheroidal cast iron

The article proposes the use of ADI cast iron as an alternative material solution for 30CrMo12 cast steel in the construction of friction systems intended for use in tribocorrosion conditions with the presence of an abradant. An independent evaluation of the resistance to tribocorrosion of the analyzed materials was made using two types of tests: a laboratory and industrial one. The industrial test was carried out using an original research stand. It made it possible to map the real operating conditions of the chain wheels of the scraper conveyor transporting a mixture of mineral abrasive and salt water. A standard tribotester with a ball-on-plate system and a model corrosion environment of 3.5% NaCl were used in the laboratory test.The research results included in the article indicate that ADI cast iron has a higher resistance to tribocorrosion than 30CrMo12 cast steel despite lower initial hardness and similar corrosion resistance. This effect may be caused by the ability to strengthen the near-surface layers of cast iron due to frictional interactions.The article also showed that the results of the comparative assessment of the tribocorrosion resistance of various materials obtained in the model systems (the laboratory test) can be as effective as the tests of real machine systems (the industrial test).

Process Design for the Wire Arc Additive Manufacturing of a Compressor Impeller

Additive manufacturing or 3D printing slowly and steadily replaces traditional subtractive manufacturing, where parts are manufactured by sequentially removing pieces of the workpiece material. In 3D printing, the material is added in layers, which enables producing parts of complex geometric shapes made from expensive metals with higher productivity and economic efficiency. In order to perform 3D printing on an industrial scale, it is necessary to combine robotics with 3D printing software. This article presents the results of the process design for 3D printing of a compressor impeller prototype. A detailed methodology of customization of the KUKA industrial robot for the 3D printing process is given. Once the impeller model was created in the CAD environment, it was cut into layers in the slicer program. Special software for offline programming of industrial robots was used to translate the program into the robot language and perform 3D printing using an experimental robotic complex. As a result, a prototype of a low-carbon steel compressor impeller was printed using a welding machine and steel wire. The developed technological process was successful and proved that an industrial robot combined with a welding machine can be used as an industrial 3D printer for printing real machine parts.

Design and development of monitoring device for corn grain cleaning loss based on piezoelectric effect

It is wildly known that the grain direct harvest can improve the corn harvest efficiency effectively. However, in the process of grain direct harvest, incomplete threshing and cleaning loss will lead to the loss of corn grain. Among them, the grain loss caused by the cleaning process can be controlled and reduced by adjusting the opening extent of the cleaning sieve, fan speed and other parameters. In traditional operation process, the operator had to adjust the cleaning mechanism based on their experience, which directly leads to low efficiency as well as high rate of grain loss. The grain loss rate sensor proposed in this paper provides the feedback parameters to control the above actuators, which makes the intelligent cleaning of the harvester possible. By comparing the existing research results, the PVDF piezoelectric film was used as the sensitive material. A sensor test platform and a real machine test platform were built and the corresponding software system of the sensor was designed and developed. The software system includes data acquisition module, filtering algorithm, wave shaped module, counting module, data processing module and wireless transmission module. Moreover, an algorithm was proposed to calibrate the error caused by misidentification of the grains and residues. After the testing and debugging, the final identification error of the laboratory tests was decreased from 12% to 3% and the error of the real machine test was under 6%, which is acceptable.The research results of this paper provided the basis for the development of intelligent control system of grain harvesting machinery, and it is of great significance to reduce the cleaning loss rate of harvesting machinery.

Rough‐set and machine learning‐based approach for optimised virtual machine utilisation in cloud computing

To meet rising demands for computing resources, information technology service providers need to select cloud-based services for their vitality and elasticity. Enormous numbers of data centres are designed to meet customer needs. Burning up energy by data centre is very high with the large-scale deployment of cloud data centres. Virtual machine consolidation strategy implementation reduces the data centre energy consumption and guarantees service level agreements. This study proposes a machine learning-based method in cloud computing for the automated use of virtual machines. Machine learning-based virtual machine selection approach integrates the migration control mechanism that enhances selection strategy efficiency. The experiment is performed with various real machine workload circumstances to provide proof and effectiveness of the proposed method. The exploratory outcome shows that the proposed approach streamlines the utilisation of the virtual machine and diminishes the consumption of energy and improves infringement of service level agreements to accomplish better performance.

Robust Chatter Stability Prediction of the Milling Process considering Uncertain Machining Positions

Milling stability is a function of the tool point frequency response functions (FRFs), which vary with the movements of the moving parts within the whole machine tool work volume. The position-dependent tool point FRFs result in uncertain prediction of the stability lobe diagram (SLD) for chatter-free machining parameter selection. Taking the variations of modal parameters to represent the variations of tool point FRFs, this paper introduces the edge theorem to predict the robust milling chatter stability. The application of the edge theorem requires the minimum and maximum modal parameters within the machining space defined by the machining position and machining allowance information. Then, radial basis function artificial neural networks (RBFANNs) are used to predict the position-dependent modal parameters in X and Y directions based on the sample information of machining positions and related modal parameters at the tool point. Moreover, sample machining spaces are determined based on the aforementioned sample positions, and the trained RBFANNs are used to obtain corresponding sample extreme modal parameters. On this basis, RBFANNs for predicting the position and machining allowance-dependent extreme modal parameters can also be trained, and they are combined with the edge theorem and zero exclusion condition to calculate robust pairs of the spindle speed (n) and limiting axial cutting depth (aplim) and then plot the robust SLD (RSLD). A case study was performed on a real three-axial vertical machining center, and the plotted RSLD considering position variations was compared with the traditional SLD. Results of the chatter tests show that the RSLD can provide more reliable (ap, n) pairs to guarantee the milling stability, validating the feasibility of the proposed robust milling chatter stability prediction method.

Machine learning to predict tamoxifen adherence among U.S. commercially insured breast cancer patients.

276 Background: Adherence to tamoxifen among women diagnosed with hormone receptor positive metastatic breast cancer (mBC) can improve survival and minimize recurrence. Screening for non-adherence at treatment initiation may support personalized care, improve health outcomes, and minimize cost of care. This study aimed to use real world data (RWD) and machine learning (ML) methods to classify tamoxifen non-adherence. Methods: A cohort of women diagnosed with incident mBC from 2012 to 2018 were identified from Truven MarketScan Commercial Claims and Encounters and Medicare supplemental administrative claims databases. Patients with < 80% proportion of days coverage (PDC) in the year following treatment initiation were classified non-adherent. Training and internal validation cohorts were randomly generated (4:1 ratio). Clinical procedures, comorbidity, treatment and healthcare encounter features in the year prior to treatment initiation were used to train logistic regression, boosted logistic regression, random forest, and feed forward neural network models and internally validated based on area under receiver operating characteristic (AUROC) curve. The most predictive ML approach was evaluated to assess feature importance. Results: A total of 3,022 patients were included with 39.9% classified as non-adherent. All ML models had moderate predictive accuracy. Logistic regression (AUROC 0.64) was easily interpreted with sensitivity 94% (95% confidence interval [CI]: 0.89, 0.92) and specificity 0.31 (95% CI: 0.29, 0.33). The model accurately classified adherence (negative predictive value 88.7%) but was non-discriminate for non-adherence (positive predictive value 47.7%). Variable importance identified top predictive factors, including patient features (≥55 years old) and pre-treatment procedures (lymphatic nuclear medicine, radiation oncology, arterial surgery). Conclusions: ML using baseline administrative data predicts tamoxifen adherence. Baseline claims may not be sufficient to predict treatment non-adherence. Further validation with enriched longitudinal data may improve model performance for incorporation of predictions into clinical decision support.

A Study to Enhance the Accuracy of Linear Motion Ball Guide System

I reported in previous paper that high precision arrangement in LMBG (Linear Motion Ball Guide) System is able to realize by introducing an art to optimize the crowning shape in both ends of raceway surface. This art was spread in industrial sector centering on Machine Tool. However, it was reported that high precision cannot be obtained with this art for the rare occasion. When I watched a number of real machines reported, it seemed that most of them were equipped with slim table. I thought that LMBG causes waving phenomena alone so that the connected table deformed infectiously, and consequently the waving amplitude did not become smaller and high precision could not be acquired. In this study, therefore, I performed an analysis which combined the load distribution theory and table deformation calculation so as to take account of table stiffness by LMBG model for the breakthrough of stiffness and waving relation. In conclusion, there found that the waving of LMBG alone transforms the table and there is a correlation between them. I led of the stiffness of LMBG and table design guideline so as to realize high accuracy including the challenge of table deformation from those relation. I report, in addition, the relation between the table stiffness and waving amplitude value were clarified through the experimental verification.

Comparative Study of Real Time Machine Learning Models for Stock Prediction through Streaming Data

Stock prediction is one of the emerging applications in the field of data science which help the companies to make better decision strategy. Machine learning models play a vital role in the field of prediction. In this paper, we have proposed various machine learning models which predicts the stock price from the real-time streaming data. Streaming data has been a potential source for real-time prediction which deals with continuous ow of data having information from various sources like social networking websites, server logs, mobile phone applications, trading oors etc. We have adopted the distributed platform, Spark to analyze the streaming data collected from two different sources as represented in two case studies in this paper. The first case study is based on stock prediction from the historical data collected from Google finance websites through NodeJs and the second one is based on the sentiment analysis of Twitter collected through Twitter API available in Stanford NLP package. Several researches have been made in developing models for stock prediction based on static data. In this work, an effort has been made to develop scalable, fault tolerant models for stock prediction from the real-time streaming data. The Proposed model is based on a distributed architecture known as Lambda architecture. The extensive comparison is made between actual and predicted output for different machine learning models. Support vector regression is found to have better accuracy as compared to other models. The historical data is considered as a ground truth data for validation.

‘Study and Research Path’ multi-approach learning of Theory of Machines and Mechanisms

ABSTRACT A Study and Research Path (SRP) is a model, based on the Anthropological Theory of the Didactic (ATD), which describes the inquiry processes that take place when a new, generating question (Q0) is considered. The investigation of the question through different study and inquiry approaches leads to a rooted tree sequence of derived questions and answers. In this work, we explore the potential of SRPs to enrich the learning of first-year’s Engineering subjects. A specific project-based SRP for the subject ‘Theory of Machines and Mechanisms' was designed, implemented and analysed, using the Didactic Engineering ATD tool. The six-month course revolved around the study of eight real machines and mechanisms found in a car, using multiple approaches (including research of information, experimentation with models, analytical and numerical simulations, discussion with peers etc). The designed SRP provided the course with a new raison d’être and helped overcoming certain problematic phenomena (algorithmisation and atomisation of problems) detected in the previous organisation. We discuss the adaptation of the students to the SRP, the changes in the learning processes based on the analysis of the media-milieu dialectics and the improvement of the subject-specific and transversal engineering skills.

A cause of natural arm-swing in bipedal walking

The research of humanoid is widely discussed whether by simulations or real machines. In human bipedal walking, swinging arms in opposite directions is a natural movement. In this research, a model of the humanoid robot, including slipping, bumping, surface-contacting and point-contacting of the foot has been established, and its dynamical equation is derived by the Newton–Euler method. And the natural arm-swing simulation has been produced, which showed that the input torque in yaw rotation of the torso could cause natural arm-swing. “Natural” means that the arm-swinging motion is induced by coupling effects existing in nonlinear dynamics of humanoid robot even though no torques have been input into shoulders. Based on the results, a hypothesis that the vibration in the yaw rotation of the torso caused natural arm swing is proposed. In this paper, we compared the arm-swing movement with or without the input torque of yaw rotation of the torso by using the above humanoid robot model. The simulation data proved the hypothesis to be valid.

Editable machine learning models? A rule-based framework for user studies of explainability

So far, most user studies dealing with comprehensibility of machine learning models have used questionnaires or surveys to acquire input from participants. In this article, we argue that compared to questionnaires, the use of an adapted version of a real machine learning interface can yield a new level of insight into what attributes make a machine learning model interpretable, and why. Also, we argue that interpretability research also needs to consider the task of humans editing the model, not least due to the existing or forthcoming legal requirements on the right of human intervention. In this article, we focus on rule models as these are directly interpretable as well as editable. We introduce an extension of the EasyMiner system for generating classification and explorative models based on association rules. The presented web-based rule editing software allows the user to perform common editing actions such as modify rule (add or remove attribute), delete rule, create new rule, or reorder rules. To observe the effect of a particular edit on predictive performance, the user can validate the rule list against a selected dataset using a scoring procedure. The system is equipped with functionality that facilitates its integration with crowdsourcing platforms commonly used to recruit participants.

OPTIMIZATION OF CNC PARAMETERS ACCORDING TO PRODUCTIVITY CRITERIA USING A MACHINE MODEL BASED ON NEURAL NETWORKS

Every machine-tool user wants to maximize the productivity of their machines looking for balance between speed, precision and lifetime of mechanical components. Nevertheless, because CNCs have wide-ranging use, their correct parametrization for each case is key to achieving the desired objectives; on the other hand, minimizing the numbers of experimental tests to be performed on the machine is essential to reduce time and costs of the set-up process. In order to solve both difficulties, this paper presents a tool to give final user necessary information to properly adjust CNC parameters according to productivity criteria. The method makes use of experimental data to obtain a model of the machine based on neural networks. With this model machining time, geometric error and smoothness of any piece to be manufactured can be predicted, and therefore minimizing test on the real machine and recommending the appropriate values for the CNC. Keywords: optimization, CNC, neural network, model, machine tool, productivity criteria.

Measuring friction at an interface using ultrasonic response.

Friction between sliding surfaces is a fundamental phenomenon prevalent in many aspects of engineering. There are many sliding contact tribometers that measure friction force in a laboratory environment. However, the transfer of laboratory data to real machine elements is unreliable. Results depend on the specimen configuration, surface condition and environment. In this work, a method has been developed that uses the nonlinear response of a high-power ultrasonic wave to deduce friction coefficient in situ at an interface. When the high-power shear wave strikes a frictional interface, relative slip can occur. It imposes a nonlinear response and causes generation of higher-order odd frequency components in received ultrasonic signals. The amplitude of the harmonics depends on contact stress and local friction coefficient. This nonlinear ultrasonic response has been investigated both numerically and experimentally. A simple one-dimensional model has been used to predict nonlinearity generation. This model has been compared with experiments conducted on aluminium rough surfaces pressed together under increasing loads. Two strategies have been used to estimate the friction coefficient by correlating experimental and numerical third-order nonlinearity. It has proved possible to determine the friction coefficient in situ at the interface; values in the range of 0.22 to 0.61 were measured for different surface configurations.

Intelligent Adaptive PID Control Using Fuzzy Broad Learning System: An Application to Tool-Grinding Servo control Systems

This paper presents an intelligent adaptive proportional-integral-derivative (PID) control method using fuzzy broad learning system (FBLS) and investigates how the method can be applied to control a tool-grinding servo control (TGSC) system. Due to accuracy, quality and geometric errors which are often difficult to capture the dynamics of the controlled plants or systems, fixed-gain PID controllers without good three-term parameters cannot meet the stringent control performance specifications of nonlinear industrial systems and servomechanisms. To accomplish better control, an adaptive PID control strategy based on the FBLS, or abbreviated as FBLS-APPID, is rigorously proposed by integrating an online parameter learning FBLS identifier together with an adaptive predictive PID control law using FBLS, to eliminate tracking error and achieve fast-tracking and disturbance rejection. Numerical simulations on the two existing discrete-time nonlinear time-delay processes are performed to show the merits and superiority of the constructed FBLS-APPID by comparing to three existing adaptive PID methods. Finally, the applicability of the proposed method is well exemplified by conducting comparatively experimental results on a servo control loop of a real TGSC machine with fixed PID gains tuned by the proposed FBLS-APPID method.

It's alive!

Scientists and engineers cross the reality gap, transferring simulated evolution into real machines.

Evaluation of an immersive virtual learning environment for operator training in mechanical and plant engineering using video analysis

AbstractA structural evaluation is imperative for developing an effective virtual learning environment. Understanding the extent to which content that has been learned virtually can be applied practically holds particular importance. A group of persons from the technical field of mechanical and plant engineering (N = 13) participated in a virtual operator training for a case application of additive manufacturing. To evaluate the virtual learning environment the participants answered quantitative questionnaires and were asked to apply what they had learned virtually to the real machine. Both the virtual training and testing phase on the real machine were recorded by video (800 minutes in total). The category system resulting from a structured qualitative video analysis with a total of 568 codes contains design‐, instruction‐ and interaction‐related optimisation potentials for further development of the virtual learning sequence. Mistakes, difficulties and other anomalies during the application on the real machine provide further revision options. The study uses video data for the first time to derive optimisation potentials and to investigate the learning transfer of virtually learned action knowledge to the real‐world activity. Practitioner NotesWhat is already known about the topic? A combination of quantitative and qualitative data for formative evaluation is feasible to improve the quality of virtual learning environments and maintain critical constructive feedback. Relevant aspects for evaluating virtual learning environments include usability, technology acceptance, motivation and learning efficiency. Different methods for measuring the learning efficiency of VR training exist, eg, subjective measures or pretests/posttests. What this paper adds: An immersive virtual environment with its specific methodical‐didactic concept is introduced. Video data are used for the first time to derive optimisation potentials in a formative evaluation. The procedure of extracting and analysing the data in a structured video analysis is introduced in the paper. Learning transfer of the virtually learned content to the real‐world context of action is assessed and applied in the field of mechanical and plant engineering on a complex real machine for the first time. Implications for practice and/or policy: The described methodical procedure using video analysis enables deriving design‐, instruction‐ and interaction‐related optimisation potentials. The results show that spatial information was transferred to the real machine without any issues, whereby the main difficulties appeared during steps that require haptic feedback. The use of IVR should be critically assessed due to the effort required in creation.

Optimization of Real-time Reverse Transcriptase Polymerase Chain Reaction for Detection of Dengue Virus

Aims: This study was set to optimize conditions for real time reverse transcriptase polymerase chain reaction (RT-PCR) for detection of dengue virus by using rapid and simple nucleic acid extraction method.
 Methodology: One step and two step real time RT-PCR were evaluated in different PCR thermocyclers. Extraction of viral RNA was done by using a simple boom method.
 Results: The real time RT-PCR technique was successfully optimized using simple and rapid method for purification of nucleic acid, ‘boom method’. The technique works better when performed in a two-step procedure and can works well with all range of real time PCR machines. The optimized real time RT-PCR used in the present study is a valuable and reliable technique for routine diagnosis of dengue. Further investigation on the cost effectiveness in adopting this technique for routine screening and monitoring of the dengue infection should be done.

Self-adjusting on-line cutting condition for high-speed milling process

The paper presents an intelligent control system for self-adjusting on-line cutting condition for high speed machining (self-HSM) with considering the tool-wear amount to keep the machined product’s quality in allowable limit. For realizing the self-HSM, the empirical analysis of variance (ANOVA) and artifical neural network (ANN) are used. The ANOVA is used for generating the empirical functions which are used as the boundary condition as well as constraint evaluation. The ANN is used for generating the new optimal cutting condition. Then, the self-HSM updates this cutting condition on the real machine — HS Super MC500. The new optimal cutting parameter is sent to the controller for updating the new machining condition to keep the machined part’s quality. The integration of the empirical analysis and ANN enables generating the optimal cutting parameters correctly and efficiently for high-speed milling.

A Systematic Literature Review on Using Machine Learning Algorithms for Software Requirements Identification on Stack Overflow

Context. The improvements made in the last couple of decades in the requirements engineering (RE) processes and methods have witnessed a rapid rise in effectively using diverse machine learning (ML) techniques to resolve several multifaceted RE issues. One such challenging issue is the effective identification and classification of the software requirements on Stack Overflow (SO) for building quality systems. The appropriateness of ML-based techniques to tackle this issue has revealed quite substantial results, much effective than those produced by the usual available natural language processing (NLP) techniques. Nonetheless, a complete, systematic, and detailed comprehension of these ML based techniques is considerably scarce. Objective. To identify or recognize and classify the kinds of ML algorithms used for software requirements identification primarily on SO. Method. This paper reports a systematic literature review (SLR) collecting empirical evidence published up to May 2020. Results. This SLR study found 2,484 published papers related to RE and SO. The data extraction process of the SLR showed that (1) Latent Dirichlet Allocation (LDA) topic modeling is among the widely used ML algorithm in the selected studies and (2) precision and recall are amongst the most commonly utilized evaluation methods for measuring the performance of these ML algorithms. Conclusion. Our SLR study revealed that while ML algorithms have phenomenal capabilities of identifying the software requirements on SO, they still are confronted with various open problems/issues that will eventually limit their practical applications and performances. Our SLR study calls for the need of close collaboration venture between the RE and ML communities/researchers to handle the open issues confronted in the development of some real world machine learning-based quality systems.

Digital Twin–oriented real-time cutting simulation for intelligent computer numerical control machining

Digital Twin has become a frontier research topic in recent years and the important development direction of intelligent manufacturing. For numerical control machining, a Digital Twin system can be used as an intelligent monitoring and analysis center by reflecting the real machining process in a virtual environment. The machining simulation is the key technology to realize this kind of application. However, existing machining simulation systems are designed for off-line situation that cannot be used directly in Digital Twin environment. The challenges for machining simulation are analyzed and explained in this article: (1) complete process data representation in simulation system; (2) executing in cooperating with computer numerical control system; (3) more efficient simulation algorithm. In order to meet these challenges, a new machining simulation system is proposed. STEP-NC standard is used to save complete process data exported from the computer-aided manufacturing system and synchronization algorithm is developed based on the communication data of computer numerical control system. Most importantly, an optimized tri-dexel-based machining simulation algorithm is developed to perform high efficiency that can follow the real machining process. Finally, a Digital Twin system for NC machining is presented that has been tested and verified in a workshop located in COMAC (Commercial Aircraft Corporation of China Ltd).