INVERSE DYNAMIC FORMULATION OF A NOVEL HYBRID MACHINE TOOL

In recent years, hybrid devices have increasingly received more research.However, few of researchers studied the dynamic analysis. The inverse dynamic analysis of a novelhybrid machine tool designed in Tsinghua University is presented. The hybrid machine tool underconsideration consists of parallel and serial structures, which is based on a new 2-DOF parallelplatform and serial orientations. The kinematics and the dynamic equations are studied first for theparallel structure through Newton-Euler approach. And then, the dynamic analysis for serialstructures is conducted. Finally, a closed-form inverse dynamic formulation is derived by using someelimination techniques. Some simulation results are also given.

Kinematic calibration method with high measurement efficiency and robust identification for hybrid machine tools

Geometric error is the main factor affecting the machining accuracy of hybrid machine tools.Kinematic calibration is an effective way to improve the geometric accuracy of hybrid machine tools.The necessity to measure both position and orientation at each pose,as well as the instability of identification in case of incomplete measurements,severely affects the application of traditional calibration methods.In this study,a kinematic calibration method with high measurement efficiency and robust identification is proposed to improve the kinematic accuracy of a five-axis hybrid machine tool.First,the configuration is introduced,and an error model is derived.Further,by investigating the mechanism error characteristics,a measurement scheme that only requires tool centre point position error measurement and one alignment operation is proposed.Subsequently,by analysing the effects of unmeasured degrees of freedom(DOFs)on other DOFs,an improved nonlinear least squares method based on virtual measurement values is proposed to achieve stable parameter identification in case of incomplete measurement,without introducing additional parameters.Finally,the proposed calibration method is verified through simulations and experiments.The proposed method can efficiently accomplish the kinematic calibration of the hybrid machine tool.The accuracy of the hybrid machine tool is significantly improved after calibration,satisfying actual aerospace machining requirements.

Human Interaction Recognition in Surveillance Videos Using Hybrid Deep Learning and Machine Learning Models

Human Interaction Recognition(HIR)was one of the challenging issues in computer vision research due to the involvement of multiple individuals and their mutual interactions within video frames generated from their movements.HIR requires more sophisticated analysis than Human Action Recognition(HAR)since HAR focuses solely on individual activities like walking or running,while HIR involves the interactions between people.This research aims to develop a robust system for recognizing five common human interactions,such as hugging,kicking,pushing,pointing,and no interaction,from video sequences using multiple cameras.In this study,a hybrid Deep Learning(DL)and Machine Learning(ML)model was employed to improve classification accuracy and generalizability.The dataset was collected in an indoor environment with four-channel cameras capturing the five types of interactions among 13 participants.The data was processed using a DL model with a fine-tuned ResNet(Residual Networks)architecture based on 2D Convolutional Neural Network(CNN)layers for feature extraction.Subsequently,machine learning models were trained and utilized for interaction classification using six commonly used ML algorithms,including SVM,KNN,RF,DT,NB,and XGBoost.The results demonstrate a high accuracy of 95.45%in classifying human interactions.The hybrid approach enabled effective learning,resulting in highly accurate performance across different interaction types.Future work will explore more complex scenarios involving multiple individuals based on the application of this architecture.

Advancing Early Detection of Colorectal Adenomatous Polyps via Genetic Data Analysis: A Hybrid Machine Learning Approach

In this study, a hybrid machine learning (HML)-based approach, incorporating Genetic data analysis (GDA), is proposed to accurately identify the presence of adenomatous colorectal polyps (ACRP) which is a crucial early detector of colorectal cancer (CRC). The present study develops a classification ensemble model based on tuned hyperparameters. Surpassing accuracy percentages of early detection approaches used in previous studies, the current method exhibits exceptional performance in identifying ACRP and diagnosing CRC, overcoming limitations of CRC traditional methods that are based on error-prone manual examination. Particularly, the method demonstrates the following CRP identification accuracy data: 97.7 ± 1.1, precision: 94.3 ± 5, recall: 96.0 ± 3, F1-score: 95.7 ± 4, specificity: 97.3 ± 1.2, average AUC: 0.97.3 ± 0.02, and average p-value: 0.0425 ± 0.07. The findings underscore the potential of this method for early detection of ACRP as well as clinical use in the development of CRC treatment planning strategies. The advantages of this approach are highly expected to contribute to the prevention and reduction of CRC mortality.

Accuracy Analysis and Calibration of Gantry Hybrid Machine Tool

The kinematic accuracy is a key factor in the design of parallel or hybrid machine tools. This analysis improved the accuracy of a 4-DOF (degree of freedom) gantry hybrid machine tool based on a 3-DOF planar parallel manipulator by compensating for various positioning errors. The machine tool architecture was described with the inverse kinematic solution. The control parameter error model was used to analyze the accuracy of the 3-DOF planar parallel manipulator and to develop a kinematic calibration method. The experimental results prove that the calibration method reduces the cutter nose errors from ±0.50 mm to ±0.03 mm for a horizontal movement of 600 mm by compensating for errors in the slider home position, the guide way distance and the extensible strut home position. The calibration method will be useful for similar types of parallel kinematic machines.

Modeling and Control of Hybrid Machine Systems—a Five-bar Mechanism Case

A hybrid machine （HM） as a typical mechatronic device, is a useful tool to generate smooth motion, and combines the motions of a large constant speed motor with a small servo motor by means of a mechnical linkage mechanism, in order to provide a powerful programmable drive system. To achieve design objectives, a control system is required. To design a better control system and analyze the performance of an HM, a dynamic model is necessary. This paper first develops a dynamic model of an HM with a five-bar mechanism using a Lagrangian formulation. Then, several important properties which are very useful in system analysis, and control system design, are presented. Based on the developed dynamic model, two control approaches, computed torque, and combined computed torque and slide mode control, are adopted to control the HM system. Simulation results demonstrate the control performance and limitations of each control approach.

Predicting the daily return direction of the stock market using hybrid machine learning algorithms

Big data analytic techniques associated with machine learning algorithms are playing an increasingly important role in various application fields,including stock market investment.However,few studies have focused on forecasting daily stock market returns,especially when using powerful machine learning techniques,such as deep neural networks(DNNs),to perform the analyses.DNNs employ various deep learning algorithms based on the combination of network structure,activation function,and model parameters,with their performance depending on the format of the data representation.This paper presents a comprehensive big data analytics process to predict the daily return direction of the SPDR S&P 500 ETF(ticker symbol:SPY)based on 60 financial and economic features.DNNs and traditional artificial neural networks(ANNs)are then deployed over the entire preprocessed but untransformed dataset,along with two datasets transformed via principal component analysis(PCA),to predict the daily direction of future stock market index returns.While controlling for overfitting,a pattern for the classification accuracy of the DNNs is detected and demonstrated as the number of the hidden layers increases gradually from 12 to 1000.Moreover,a set of hypothesis testing procedures are implemented on the classification,and the simulation results show that the DNNs using two PCA-represented datasets give significantly higher classification accuracy than those using the entire untransformed dataset,as well as several other hybrid machine learning algorithms.In addition,the trading strategies guided by the DNN classification process based on PCA-represented data perform slightly better than the others tested,including in a comparison against two standard benchmarks.

A novel hybrid estimation of distribution algorithm for solving hybrid flowshop scheduling problem with unrelated parallel machine

The hybrid flow shop scheduling problem with unrelated parallel machine is a typical NP-hard combinatorial optimization problem, and it exists widely in chemical, manufacturing and pharmaceutical industry. In this work, a novel mathematic model for the hybrid flow shop scheduling problem with unrelated parallel machine(HFSPUPM) was proposed. Additionally, an effective hybrid estimation of distribution algorithm was proposed to solve the HFSPUPM, taking advantage of the features in the mathematic model. In the optimization algorithm, a new individual representation method was adopted. The(EDA) structure was used for global search while the teaching learning based optimization(TLBO) strategy was used for local search. Based on the structure of the HFSPUPM, this work presents a series of discrete operations. Simulation results show the effectiveness of the proposed hybrid algorithm compared with other algorithms.

Hybrid Excited Permanent Magnet Machines for Electric and Hybrid Electric Vehicles

This paper reviews various hybrid excited(HE)machines from the perspective of location of PM and DC excitation,series/parallel connection of PM and DC excited magnetic fields,and 2D/3D magnetic fields,respectively.The advantages as well as drawbacks of each category are analyzed.Since an additional control degree,i.e.DC excitation,is introduced in the HE machine,the flux weakening control strategies are more complex.The flux weakening performance as well as efficiency are compared with different control strategies.Then,the potential to mitigate the risk of uncontrolled overvoltage fault at high speed operation is highlighted by controlling the field excitation.Since additional DC coils are usually required for HE machines compared with pure PM excitation,the spatial confliction inevitably results in electromagnetic performance reduction.Finally,the technique to integrate the field and armature windings with open-winding drive circuit is introduced,and novel HE machines without a DC coil are summarized.

Study on Applying Hybrid Machine Learning into Family Apparel Expenditure

Hybrid Machine Learning (HML) is a kind of advanced algorithm in the field of intelligent information process. It combines the induced learning based-on decision-making tree with the blocking neural network. And it provides a useful intelligent knowledge-based data mining technique. Its core algorithm is ID3 and Field Theory based ART (FTART). The paper introduces the principals of hybrid machine learning firstly, and then applies it into analyzing family apparel expenditures and their influencing factors systematically. Finally, compared with those from the traditional statistic methods, the results from HML is more friendly and easily to be understood. Besides, the forecasting by HML is more correct than by the traditional ways.

Research on the IL-Bagging-DHKELM Short-Term Wind Power Prediction Algorithm Based on Error AP Clustering Analysis

With the continuous advancement of China’s“peak carbon dioxide emissions and Carbon Neutrality”process,the proportion of wind power is increasing.In the current research,aiming at the problem that the forecasting model is outdated due to the continuous updating of wind power data,a short-term wind power forecasting algorithm based on Incremental Learning-Bagging Deep Hybrid Kernel Extreme Learning Machine(IL-Bagging-DHKELM)error affinity propagation cluster analysis is proposed.The algorithm effectively combines deep hybrid kernel extreme learning machine(DHKELM)with incremental learning(IL).Firstly,an initial wind power prediction model is trained using the Bagging-DHKELM model.Secondly,Euclidean morphological distance affinity propagation AP clustering algorithm is used to cluster and analyze the prediction error of wind power obtained from the initial training model.Finally,the correlation between wind power prediction errors and Numerical Weather Prediction(NWP)data is introduced as incremental updates to the initial wind power prediction model.During the incremental learning process,multiple error performance indicators are used to measure the overall model performance,thereby enabling incremental updates of wind power models.Practical examples show the method proposed in this article reduces the root mean square error of the initial model by 1.9 percentage points,indicating that this method can be better adapted to the current scenario of the continuous increase in wind power penetration rate.The accuracy and precision of wind power generation prediction are effectively improved through the method.

Design and Analysis of New Modular Stator Hybrid Excitation Synchronous Motor

Hybrid excitation motor is a combination of permanent magnet motor and electric excitation motor,which can flexibly adjust the air gap magnetic field.At present,the traditional silicon steel sheet core material is widely used,but this material limits the further reduction of stator iron loss.In this paper,a new type of hybrid excitation synchronous motor with modular stator structure based on amorphous alloy material is proposed.The design power is 1kW,and the speed is 3000rpm.By placing the armature winding and electric excitation winding in the stator slot,the slip ring and brush are avoided,and the reliability of the motor is improved.The rotor adopts staggered magnetic pole structure,which has strong flux adjusting ability.The core loss is greatly reduced by using amorphous alloy.Firstly,the structure and working principle of the new motor are given;Secondly,the size parameters of the motor are given,and the principle of flux adjustment is verified and analyzed by three-dimensional finite element(3D-FEM);Finally,through theoretical analysis of the influence factors of the magnetic adjustment ability and 3D-FEM finite element computation,the flux adjustment ability and the torque lifting at low speed are verified,and the advantages of the motor are verified.

Identifying Severity of COVID-19 Medical Images by Categorizing Using HSDC Model

Since COVID-19 infections are increasing all over the world,there is a need for developing solutions for its early and accurate diagnosis is a must.Detectionmethods for COVID-19 include screeningmethods like Chest X-rays and Computed Tomography(CT)scans.More work must be done on preprocessing the datasets,such as eliminating the diaphragm portions,enhancing the image intensity,and minimizing noise.In addition to the detection of COVID-19,the severity of the infection needs to be estimated.The HSDC model is proposed to solve these problems,which will detect and classify the severity of COVID-19 from X-ray and CT-scan images.For CT-scan images,the histogram threshold of the input image is adaptively determined using the ICH Swarm Optimization Segmentation(ICHSeg)algorithm.Based on the Statistical and Shape-based feature vectors(FVs),the extracted regions are classified using a Hybrid model for CT images(HSDCCT)algorithm.When the infections are detected,it’s classified as Normal,Moderate,and Severe.A fused FHI is formed for X-ray images by extracting the features of Histogram-oriented gradient(HOG)and Image profile(IP).The FHI features of X-ray images are classified using Hybrid Support Vector Machine(SVM)and Deep Convolutional Neural Network(DCNN)HSDCX algorithm into COVID-19 or else Pneumonia,or Normal.Experimental results have shown that the accuracy of the HSDC model attains the highest of 94.6 for CT-scan images and 95.6 for X-ray images when compared to SVM and DCNN.This study thus significantly helps medical professionals and doctors diagnose COVID-19 infections quickly,which is the most needed in current years.

Estimating SVCV waterborne transmission and predicting experimental epidemic development:A modeling study using a machine learning approach

Viral infectious diseases significantly threaten the sustainability of freshwater fish aquaculture.The lack of studies on epidemic transmission patterns and mechanisms inhibits the development of containment strategies from the viewpoint of veterinary public health.This study raises an epidemic mathematical model considering water transmission with the aim of analyzing the transmission process more accurately.The basic reproduction number R0 was derived by the model parameter including the water transmission coefficient and was used for the analysis of the virus transmission.Spring viremia of carp virus(SVCV)and zebrafish were used as model viruses and animals,respectively,to conduct the transmission experiment.Transmission through water was achieved by connecting two aquarium tanks with a water channel but blocking the fish movement between the tanks.With the collected experimental data,we determined the optimal hybrid machine learning algorithm to analyze the transmission process using an established mathematical model.In addition,future transmission was predicted and validated using the epidemic model and an optimal algorithm.Finally,the sensitivity of model parameters and the simulations of R0 variation were performed based on the modified complex epidemic model.This study is of significance in providing theoretical guidance for minimizing R0 by manipulating model parameters with containment measures.More importantly,since the modified model and algorithm demonstrated better performance in handling freshwater fish transmission problems,this study advances the future application of transmissible disease modeling with larger datasets in freshwater fish aquaculture.

Dynamic Feedforward Control of a Novel 3-PSP 3-DOF Parallel Manipulator

Parallel manipulators with less than six degrees of freedom （DOF） have been increasingly used in high-speed hybrid machine tools. The structural features of parallel manipulators are dynamic, a characteristic that is particularly significant when these manipulators are used in high-speed machine tools. However, normal kinematic control method cannot satisfy the requirements of the control system. Many researchers use model-based dynamic control methods, such as the dynamic feedforward control method. However, these methods are rarely used in hybrid machine tools because of the complex dynamic model of the parallel manipulator. In order to study the dynamic control method of parallel manipulators, the dynamic feedforward control method is used in the dynamic control system of a 3-PSP （prismatic-spherical-prismatic） 3-DOF spatial parallel manipulator used as a spindle head in a high-speed hybrid machine tool. Using kinematic analysis as basis and the Newton-Euler method, we derive the dynamic model of the parallel manipulator. Furthermore, a model-based dynamic feedforward control system consisting of both kinematic control and dynamic control subsystems is established. The dynamic control subsystem consists of two modules. One is used to eliminate the influence of the dynamic characteristics of high-speed movement, and the other is used to eliminate the dynamic disturbances in the milling process. Finally, the simulation model of the dynamic feedforward control system of the 3-PSP parallel manipulator is constructed in Matlab/Simulink. The simulations of the control system eliminating the influence of the dynamic characteristics and dynamic disturbances are conducted. A comparative study between the simulations and the normal kinematic control method is also presented.The simulations prove that the dynamic feedforward control method effectively eliminates the influence of the dynamic disturbances and dynamic characteristics of the parallel manipulator on high-speed machine tools, and significantly improves the trajectory accuracy. This is the first attempt to introduce the dynamic feedfordward control method into the 3-PSP spatial parallel manipulator whose dynamic model is complex and provides a study basis for the real-time dynamic control of the high-speed hybrid machine tools.

Compressive strength prediction of sprayed concrete lining in tunnel engineering using hybrid machine learning techniques

Sprayed concrete lining is a commonly employed support measure in tunnel engineering,which plays an important role in construction safety.Compressive strength is a key performance indicator of sprayed concrete lining,and the traditional measuring method is time-consuming and laborious.This paper proposes various hybrid machine learning algorithms to accomplish the advanced prediction of compressive strength of sprayed concrete lining based on the mixture design.Two hundred and five sets of experimental data were collected from a water conveyance tunnel in northwestern China for model construction,and each set of data was made up of six basic input variables(i.e.,water,cement,mineral powder,superplasticizer,coarse aggregate,and fine aggregate)and one output variable(i.e.,compressive strength).In order to eliminate the correlation between input variables,a new composite indicator(i.e.,the water-binder ratio)was introduced to achieve dimensionality reduction.After that,four hybrid models in total were built,namely BPNN-QPSO,SVR-QPSO,ELM-QPSO,and RF-QPSO,where the hyper-parameters of BPNN,SVR,ELM,and RF were auto-tuned by QPSO.Engineering application results indicated that RF-QPSO achieved the lowest mean absolute percentage error(MAPE)of 3.47% and root mean square error(RMSE)of 1.30 and the highest determination coefficient(R^(2))of 0.93 in the four hybrid models.Moreover,RFQPSO had the shortest running time of 0.15 s,followed by SVR-QPSO(0.18 s),ELM-QPSO(1.19 s),and BPNN-QPSO(1.58 s).Compared with BPNN-QPSO,SVR-QPSO,and ELM-QPSO,RF-QPSO performed the most superior performance in terms of both prediction accuracy and running speed.Finally,the importance of input variables on the model performance was quantitatively evaluated,further enhancing the interpretability of RF-QPSO.

An Analytical Method of Calculating Back-EMF in Dual Consequent Hybrid Excitation Synchronous Machine

In order to solve the problem of asymmetric bidirectional flux control capability in hybrid excitation machine,a novel structure called dual consequent hybrid excitation synchronous(DCHES)machine is presented in this paper.Generally,the analysis of back-EMF for the machine with complex electromagnetic structure such as DCHES machine should utilize 3-D finite element analysis(FEA),which will require huge resources and computing time.In order to avoid using 3-D FEA to analyze the back-EMF of complex structure,an analytical method of calculating back-EMF is presented in this paper.The electromagnetic field in 3-D space can be simplified as a 2-D field by dividing the 3-D field into several simple zones,the resultant effect equals to the summation of every single 2-D field's effect.According to electromagnetic theory,the analytical formula of back-EMF is obtained on the basis of Fourier series.The influence of main parameters on back-EMF waveform under sine and trapezoidal flux distribution is discussed respectively.The theoretical result shows that the trapezoidal air-gap flux distribution would generate a sine back-EMF.Finally,the presented analytical method is verified and evaluated with experimental results.

Multi-objective optimisation with hybrid machine learning strategy for complex catalytic processes

Catalytic chemical processes such as hydrocracking,gasification and pyrolysis play a vital role in the renewable energy and net zero transition.Due to the complex and non-linear behaviours during operation,catalytic chemical processes require a powerful modelling tool for prediction and optimisation for smart operation,speedy green process routes discovery and rapid process design.However,challenges remain due to the lack of an effective modelling and optimisation toolbox,which requires not only a precise analysis but also a fast optimisation.Here,we propose a hybrid machine learning strategy by embedding the physics-based continuum lumping kinetic model into the data-driven artificial neural network framework.This hybrid model is adopted as the surrogate model in the multi-objective optimisation and demonstrated in the benchmarking of a hydrocracking process.The results show that the novel hybrid surrogate model exhibits the mean square error less than 0.01 by comparing with the physics-based simulation results.This well-trained hybrid model was then integrated with non-dominated-sort genetic algorithm(NSGA-II)as the surrogate model to evaluate and optimise the yield and selectivity of the hydrocracking process.The Pareto front from the multi-objective optimisation was able to identify the trade-off curve between the objective functions which is essential for the decision-making during process design.Our work indicates that adopting the hybrid machine learning strategy as the surrogate model in the multi-objective optimisation is a promising approach in various complex catalytic chemical processes to enable an accurate computation as well as a rapid optimisation.

Base placement optimization of a mobile hybrid machining robot by stiffness analysis considering reachability and nonsingularity constraints

The mobile hybrid machining robot has a very bright application prospect in the field of high-efficiency and high-precision machining of large aerospace structures.However,an inappropriate base placement may make the robot encounter a singular configuration,or even fail to complete the entire machining task due to unreachability.In addition to considering the two constraints of reachability and non-singularity,this paper also optimizes the robot base placement with stiffness as the goal to improve the machining quality.First of all,starting from the structure of the robot,the reachability and nonsingularity constraints are transformed into a simple geometric constraint imposed on the base placement:feasible base placement area.Then,genetic algorithm is used to search for the base placement with near optimal stiffness(near optimal base placement for short)in the feasible base placement area.Finally,multiple controlled experiments were carried out by taking the milling of a protuberance on the spacecraft cabin as an example.It is found that the calculated optimal base placement meets all the constraints and that the machining quality was indeed improved.In addition,compared with simple genetic algorithm,it is proved that the feasible base placement area method can shorten the running time of the whole program.

Surface-improvement mechanism of hybrid electrochemical discharge process using variable-amplitude pulses

Superalloys are commonly used in aircraft manufacturing;however,the requirements for high surface quality and machining accuracy make them difficult to machine.In this study,a hybrid electrochemical discharge process using variable-amplitude pulses is proposed to achieve this target.In this method,electrochemical machining(ECM)and electrical discharge machining(EDM)are unified into a single process using a sequence of variable-amplitude pulses such that the machining process realizes both good surface finish and high machining accuracy.Furthermore,the machining mechanism of the hybrid electrochemical discharge process using variable-amplitude pulses is studied.The mechanism is investigated by observations of machining waveforms and machined surface.It is found that,with a high-frequency transformation between high-and low-voltage waveforms within a voltage cycle,the machining mechanism is frequently transformed from EDM to pure ECM.The critical discharge voltage is 40 V.When pulse voltages greater than 40 V are applied,the machining accuracy is good;however,the surface has defects such as numerous discharge craters.High machining accuracy is maintained when high-voltage pulses are replaced by low-voltage pulses to enhance electrochemical dissolution.The results indicate that the proposed hybrid electrochemical discharge process using variable-amplitude pulses can yield high-quality surfaces with high machining accuracy.