MENDED GENETIC BP NETWORK AND APPLICATION TO ROLLING FORCE PREDICTION OF 4-STAND TANDEM COLD STRIP MILL

In order to make good use of the ability to approach any function of BP (backpropagation) network and overcome its local astringency, and also make good use of the overallsearch ability of GA (genetic algorithms), a proposal to regulate the network's weights using bothGA and BP algorithms is suggested. An integrated network system of MGA (mended genetic algorithms)and BP algorithms has been established. The MGA-BP network's functions consist of optimizing GAperformance parameters, the network's structural parameters, performance parameters, and regulatingthe network's weights using both GA and BP algorithms. Rolling forces of 4-stand tandem cold stripmill are predicted by the MGA-BP network, and good results are obtained.

Prediction of Grinding Force by an Electroplated Grinding Wheel with Orderly-Micro-Grooves

The ability to predict a grinding force is important to control,monitor,and optimize the grinding process.Few theoretical models were developed to predict grinding forces when a structured wheel was used in a grinding process.This paper aimed to establish a single-grit cutting force model to predict the ploughing,friction and cutting forces in a grinding process.It took into the consideration of actual topography of the grinding wheel,and a theoretical grinding force model for grinding hardened AISI 52100 by the wheel with orderly-micro-grooves was proposed.The model was innovative in the sense that it represented the random thickness of undeformed chips by a probabilistic expression,and it reflected the microstructure characteristics of the structured wheel explicitly.Note that the microstructure depended on the randomness of the protruding heights and distribution density of the grits over the wheel.The proposed force prediction model was validated by surface grinding experiments,and the results showed(1)a good agreement of the predicted and measured forces and(2)a good agreement of the changes of the grinding forces along with the changes of grinding parameters in the prediction model and experiments.This research proposed a theoretical grinding force model of an electroplated grinding wheel with orderly-micro-grooves which is accurate,reliable and effective in predicting grinding forces.

Prediction of Dynamic Cutting Force and Regenerative Chatter Stability in Inserted Cutters Milling

Currently, the modeling of cutting process mainly focuses on two aspects: one is the setup of the universal cutting force model that can be adapted to a broader cutting condition; the other is the setup of the exact cutting force model that can accurately reflect a true cutting process. However, there is little research on the prediction of chatter stablity in milling. Based on the generalized mathematical model of inserted cutters introduced by ENGIN, an improved geometrical, mechanical and dynamic model for the vast variety of inserted cutters widely used in engineering applications is presented, in which the average directional cutting force coefficients are obtained by means of a numerical approach, thus leading to an analytical determination of stability lobes diagram (SLD) on the axial depth of cut. A new kind of SLD on the radial depth of cut is also created to satisfy the special requirement of inserted cutter milling. The corresponding algorithms used for predicting cutting forces, vibrations, dimensional surface finish and stability lobes in inserted cutter milling under different cutting conditions are put forward. Thereafter, a dynamic simulation module of inserted cutter milling is implemented by using hybrid program of Matlab with Visual Basic. Verification tests are conducted on a vertical machine center for Aluminum alloy LC4 by using two different types of inserted cutters, and the effectiveness of the model and the algorithm is verified by the good agreement of simulation result with that of cutting tests under different cutting conditions. The proposed model can predict the cutting process accurately under a variety of cutting conditions, and a high efficient and chatter-free milling operation can be achieved by a cutting condition optimization in industry applications.

IMPROVED MODEL FOR THREE DIMENSIONAL NONLINEAR WATER WAVE FORCE PREDICTION

An improved model for numerically predicting nonlinear wave forces exerted on an offshore structure is pro- posed.In a previous work[9],the authors presented a model for the same purpose with an open boundary condi- tion imposed,where the wave celerity has been defined constant.Generally,the value of wave celerity is time-de- pendent and varying with spatial location.With the present model the wave celerity is evaluated by an upwind dif- ference scheme,which enables the method to be extended to conditions of variable finite water depth,where the value of wave celerity varies with time as the wave approaches the offshore structure.The finite difference method incorporated with the time-stepping technique in time domain developed here makes the numerical evolution effec- tive and stable.Computational examples on interactions between a surface-piercing vertical cylinder and a solitary wave or a cnoidal wave train demonstrates the validity of this program.

Force Feedback Coupling with Dynamics for Physical Simulation of Product Assembly and Operation Performance

Most existing force feedback methods are still difficult to meet the requirements of real-time force calculation in virtual assembly and operation with complex objects. In addition, there is often an assumption that the controlled objects are completely flee and the target object is only completely fixed or flee, thus, the dynamics of the kinematic chain where the controlled objects are located are neglected during the physical simulation of the product manipulation with force feedback interaction. This paper proposes a physical simulation method of product assembly and operation manipulation based on statistically learned contact force prediction model and the coupling of force feedback and dynamics. In the proposed method, based on hidden Markov model （HMM） and local weighting learning （LWL）, contact force prediction model is constructed, which can estimate the contact force in real time during interaction. Based on computational load balance model, the computing resources are dynamically assigned and the dynamics integral step is optimized. In addition, smoothing process is performed to the force feedback on the synchronization points. Consequently, we can solve the coupling and synchronization problems of high-frequency feedback force servo. low-frequency dynamics solver servo and scene rendering servo, and realize highly stable and accurate force feedback in the physical simulation of product assembly and operation manipulation. This research proposes a physical simulation method of product assembly and operation manipulation.

The prediction of external flow field and hydrodynamic force with limited data using deep neural network

Predicting the external flow field with limited data or limited measurements has attracted long-time interests of researchers in many industrial applications.Physics informed neural network(PINN)provides a seamless framework for combining the measured data with the deep neural network,making the neural network capable of executing certain physical constraints.Unlike the data-driven model to learn the end-to-end mapping between the sensor data and high-dimensional flow field,PINN need no prior high-dimensional field as the training dataset and can construct the mapping from sensor data to high dimensional flow field directly.However,the extrapolation of the flow field in the temporal direction is limited due to the lack of training data.Therefore,we apply the long short-term memory(LSTM)network and physics-informed neural network(PINN)to predict the flow field and hydrodynamic force in the future temporal domain with limited data measured in the spatial domain.The physical constraints(conservation laws of fluid flow,e.g.,Navier-Stokes equations)are embedded into the loss function to enforce the trained neural network to capture some latent physical relation between the output fluid parameters and input tempo-spatial parameters.The sparsely measured points in this work are obtained from computational fluid dynamics(CFD)solver based on the local radial basis function(RBF)method.Different numbers of spatial measured points(4–35)downstream the cylinder are trained with/without the prior knowledge of Reynolds number to validate the availability and accuracy of the proposed approach.More practical applications of flow field prediction can compute the drag and lift force along with the cylinder,while different geometry shapes are taken into account.By comparing the flow field reconstruction and force prediction with CFD results,the proposed approach produces a comparable level of accuracy while significantly fewer data in the spatial domain is needed.The numerical results demonstrate that the proposed approach with a specific deep neural network configuration is of great potential for emerging cases where the measured data are often limited.

Forces in Hard Turning of 51CrV4 with Wiper Cutting Tool

For precision machining, the hard turning process is becoming an important alternative to some of the existing grinding processes. This paper presents an analytical model for predicting cutting forces in hard turning of 51CRV4 with hardness of 68 HRC. The cutting tool used is made from cubic boron nitride （CBN） with a wiper cutting edge. Formulas for differential chip loads are derived for three different situations, depending on the radial depth of cut. The cutting forces are determined by integrating the differential cutting forces over the tool-workpiece engagement domain. For validation, cutting forces predicted by the model were compared with experimental measurements, and most of the results agree quite well.

Study on the Predictive of Dynamic Milling Force of Milling Process Based on Data Mining

To improve surface accuracy of the work-piece and obtain potentially valuable information,a dynamic milling force prediction model was proposed based on data mining.In view of the current dynamic milling force obtained through finite element simulation and analytical calculation,in the finite element modeling,the model built is inevitably different from the actual working conditions,and the analytical calculation is slightly cumbersome and complex,and a dynamic milling force prediction model based on data mining is proposed.The model was established using a combination of regression analysis and Radial Basis Function(RBF) neural network.Using data mining as a means,the internal relationship between milling force,cutting parameters,temperature,vibration and surface quality is deeply analyzed,and the influence of dynamic milling force changes on different situations is extracted and summarized by the methods of cluster analysis and correlation analysis.The results show that the proposed dynamic milling force model has a good prediction effect,ensures the production quality,reduces the occurrence of flutter,improves the surface accuracy of the work-piece,and provides a more accurate basis for the selection of process parameters.

A dynamic tire model based on HPSO-SVM

In order to accurately describe the force mechanism of tires on agricultural roads and improve the life cycle of agricultural tires,a tire-deformable terrain model was established.The effects of tread pattern,wheel spine,tire sidewall elasticity,inflation pressure and soil deformation were considered in the model and fitted with a support vector machine(SVM)model.Hybrid particle swarm optimization(HPSO)was used to optimize the parameters of SVM prediction model,of which inertia weight and learning factor were improved.To verify the performance of the model,a tire force prediction model of agricultural vehicle with the improved SVM method was investigated,which was a complex nonlinear problem affected by many factors.Cross validation(CV)method was used to evaluate the training precision accuracy of the model,and then the improved HPSO was adopted to select parameters.Results showed that the choice randomness of specifying the parameters was avoided and the workload of the parameter selection was reduced.Compared with the dynamic tire model without considering the influence of tread pattern and wheel spine,the improved SVM model achieved a better prediction performance.The empirical results indicate that the HPSO based parameters optimization in SVM is feasible,which provides a practical guidance to tire force prediction of agricultural transport vehicles.

Validation of surface coating with nanoparticles to improve the flowability of fine cohesive powders

Fluidization of fine cohesive powders is seriously restricted by the strong interparticle cohesion. The rational combination of nanoparticles with fine cohesive powders is expected to obtain composite par- ticles with improved flowability. In this work, we firstly reviewed the sandwich and three-point contact models regarding the fundamental principles of nano-additives in reducing cohesiveness. Based on these previous models, the effects of the size of nanoparticles, their agglomeration and coverage on the surface of cohesive powders in reducing interparticle forces were theoretically analyzed. To validate the the- ory effectiveness for the irregularly shaped cohesive powders, an extreme case of cubic powders coated with silica nanoparticles was fabricated, and the flowability of the composite particles was determined experimentally. Ultimately, based oN force balance of a single particle, a semi-theoretical criterion for predicting the fluidization behavior of coated powders was developed to guide the practical applications of improving the flowability of cohesive powders through structural design and modulation.