Wyniki wyszukiwania - BazTech

1

Diagnosis of Missed Ductile Iron Melts with Process Modelling

Perzyk M., Werlaty M.

Archives of Foundry Engineering

|

2017

|

Vol. 17, iss. 4

123--126

EN

The paper presents an application of advanced data-driven (soft) models in finding the most probable particular causes of missed ductile iron melts. The proposed methodology was tested using real foundry data set containing 1020 records with contents of 9 chemical elements in the iron as the process input variables and the ductile iron grade as the output. This dependent variable was of discrete (nominal) type with four possible values: ‘400/18’, ‘500/07’, ‘500/07 special’ and ‘non-classified’, i.e. the missed melt. Several types of classification models were built and tested: MLP-type Artificial Neural Network, Support Vector Machine and two versions of Classification Trees. The best accuracy of predictions was achieved by one of the Classification Tree model, which was then used in the simulations leading to conversion of the missed melts to the expected grades. Two strategies of changing the input values (chemical composition) were tried: content of a single element at a time and simultaneous changes of a selected pair of elements. It was found that in the vast majority of the missed melts the changes of single elements concentrations have led to the change from the non-classified iron to its expected grade. In the case of the three remaining melts the simultaneous changes of pairs of the elements’ concentrations appeared to be successful and that those cases were in agreement with foundry staff expertise. It is concluded that utilizing an advanced data-driven process model can significantly facilitate diagnosis of defective products and out-of-control foundry processes.

2

Methodology of Fault Diagnosis in Ductile Iron Melting Process

Perzyk M., Kozlowski J.

Archives of Foundry Engineering

|

2016

|

Vol. 16, iss. 4

101--108

EN

Statistical Process Control (SPC) based on the Shewhart’s type control charts, is widely used in contemporary manufacturing industry, including many foundries. The main steps include process monitoring, detection the out-of-control signals, identification and removal of their causes. Finding the root causes of the process faults is often a difficult task and can be supported by various tools, including data-driven mathematical models. In the present paper a novel approach to statistical control of ductile iron melting process is proposed. It is aimed at development of methodologies suitable for effective finding the causes of the out-of-control signals in the process outputs, defined as ultimate tensile strength (Rm) and elongation (A5), based mainly on chemical composition of the alloy. The methodologies are tested and presented using several real foundry data sets. First, correlations between standard abnormal output patterns (i.e. out-of-control signals) and corresponding inputs patterns are found, basing on the detection of similar patterns and similar shapes of the run charts of the chemical elements contents. It was found that in a significant number of cases there was no clear indication of the correlation, which can be attributed either to the complex, simultaneous action of several chemical elements or to the causes related to other process variables, including melting, inoculation, spheroidization and pouring parameters as well as the human errors. A conception of the methodology based on simulation of the process using advanced input - output regression modelling is presented. The preliminary tests have showed that it can be a useful tool in the process control and is worth further development. The results obtained in the present study may not only be applied to the ductile iron process but they can be also utilized in statistical quality control of a wide range of different discrete processes.

3

Adjustment method of parameters intended for first-principle models

Czop P., Kost G., Sławik D., Wszołek G., Jakubowski D.

Journal of Achievements in Materials and Manufacturing Engineering

|

2012

|

Vol. 55, nr 2

446--453

EN

Purpose: This paper demonstrates a process of estimation phenomenological parameters of a first-principle nonlinear model based on the hydraulic damper system. Design/methodology/approach: First-principle (FP) models are formulated using a system of continuous ordinary differential equations capturing usually nonlinear relations among variables of the model. The considering model applies three categories of parameters: geometrical, physical and phenomenological. Geometrical and physical parameters are deduced from construction or operational documentation. The phenomenological parameters are the adjustable ones, which are estimated or adjusted based on their roughly known values, e.g. friction/damping coefficients. Findings: A phenomenological parameter, friction coefficient, was successfully estimated based on the experimental data. The error between the model response and experimental data is not greater than 10%. Research limitations/implications: Adjusting a model to data is, in most cases, a non-convex optimization problem and the criterion function may have several local minima. This is a case when multiple parameters are simultaneously estimated. Practical implications: First-principle models are fundamental tools for understanding, optimizing, designing, and diagnosing technical systems since they are updatable using operational measurements. Originality/value: First-principle models are frequently adjusted by trial-and-error, which can lead to nonoptimal results. In order to avoid deficiencies of the trial-and-error approach, a formalized mathematical method using optimization techniques to minimize the error criterion, and find optimal values of tunable model parameters, was proposed and demonstrated in this work.

4

Demonstration of first-principle model adjustment approach with the use of a simplified heater model

Barszcz T., Bednarz J., Czop P.

Diagnostyka

|

2012

|

nr 1(61)

39-44

EN

The paper discusses a process of formulation and identification of a first-principle data-driven heater model. The model is formulated using a system of continuous ordinary differential equations capturing usually nonlinear relations among variables of the model. The considering model applies three categories of parameters: geometrical, physical and phenomenological. Geometrical and physical parameters are deduced from construction or operational documentation. The phenomenological parameters are the adjustable ones. First-principle models are frequently adjusted by trial-and-error, which can lead to non-optimal results. In order to avoid deficiencies of the trial-and-error approach, a formalized mathematical method using optimization techniques to minimize the error criterion, and find optimal values of adjustable model parameters, was proposed and demonstrated in this work.

PL

Artykuł omawia proces modelowania podgrzewacza regeneracyjnego pracującego w systemie bloku energetycznego z wykorzystaniem strojonych równań fizycznych. Model jest formułowany z użyciem układu zwyczajnych równań różniczkowych obejmujących wzajemne nieliniowe relacje pomiędzy zmiennymi modelu. Rozważany model stosuje trzy kategorie parametrów: geometryczne, fizyczne, oraz fenomenologiczne. Parametry geometryczne oraz fizyczne są ustalane na podstawie dokumentacji konstrukcyjnej oraz operacyjnej. Parametrami strojonym są parametry fenomenologiczne. Modele wyprowadzane na podstawie praw fizycznych są często strojone metodą prób i błędów, co może prowadzić do nieoptymalnych wyników. Dla ominięcia wad metody została zastosowana metoda najmniejszych kwadratów do strojenia parametrów fenomenologicznych modelu podgrzewacza tj. współczynników wymiany ciepła.

5

Formulation and identification of First-Principle Data-Driven models

Czop P., Kost G., Sławik D., Wszołek G.

Journal of Achievements in Materials and Manufacturing Engineering

|

2011

|

Vol. 44, nr 2

179--186

EN

Purpose The paper consists of two parts. The first part presents and discusses a process of formulation and identification of First-Principle Data-Driven (FPDD) models, while the second part demonstrates numerical examples of identification of FPDD models. Design/methodology/approach: First-Principle (FP) model is formulated using a system of continuous ordinary differential equations capturing usually nonlinear relations among variables of the model. The considering model applies three categories of parameters: geometrical, physical and phenomenological. Geometrical and physical parameters are deduced from construction or operational documentation. The phenomenological parameters are the adjustable ones, which are estimated or adjusted based on their roughly known values, e.g. friction/damping coefficients. Findings A few phenomenological parameters were successfully estimated from numerically generated data. The error between the true and estimated value of the parameter occurred, however its magnitude is low at level below 2%. Research limitations/implications Adjusting a model to data is, in most cases, a non-convex optimization problem and the criterion function may have several local minima. This is a case when multiple parameters are simultaneously estimated. Practical implications: FPDD models are an excellent tool for understanding, optimizing, designing, and diagnosing technical systems since they are updatable using operational measurements. This opens application area, for example, for model-based design and early warning diagnostics. Originality/value: First-Principle (FP) models are frequently adjusted by trial-and-error, which can lead to non-optimal results. In order to avoid deficiencies of the trial-and-error approach, a formalized mathematical method using optimization techniques to minimize the error criterion, and find optimal values of tunable model parameters, was proposed and demonstrated in this work.