Purpose

This paper presents the application of grey modeling for thermal spray processing parameter analysis in less data environment.

Design/methodology/approach

Based on processing knowledge, key processing parameters of thermal spray process are analyzed and preselected. Then, linear and non-linear grey modeling models are integrated to mine the relationships between different processing parameters.

Findings

Model A reveals the linear correlation between the HVOF process parameters and the characterization of particle in-flight with average relative errors of 9.230 percent and 5.483 percent for velocity and temperature.

Research limitations/implications

The prediction accuracies of coatings properties vary, which means that there exists more complex non-linear relationship between the identified input parameters and coating results, or more unexpected factors (e.g. factors from material side) should be further investigated.

Practical implications

According to the modeling case in this paper, method has potential to deal with other diverse modeling problems in different industrial applications where challenge to collecting large quantity of data sets exists.

Originality/value

It is the first time to apply grey modeling for thermal spray processing where complicated relationships among processing parameters exist. The modeling results show reasonable results to experiment and existing processing knowledge.

中文翻译：

---

灰色模型用于热喷涂工艺参数分析

目的

本文介绍了灰色模型在较少数据环境下用于热喷涂工艺参数分析的应用。

设计/方法/方法

根据加工知识，对热喷涂工艺的关键加工参数进行分析和预选。然后，整合线性和非线性灰色建模模型以挖掘不同处理参数之间的关系。

发现

模型A揭示了HVOF工艺参数与飞行中颗粒表征之间的线性相关性，速度和温度的平均相对误差为9.230％和5.483％。

研究局限/意义

涂层性能的预测准确性会有所不同，这意味着在识别出的输入参数和涂层结果之间存在更复杂的非线性关系，或者应进一步研究更多意外因素（例如，来自材料方面的因素）。

实际影响

根据本文中的建模案例，该方法具有解决不同工业应用中其他各种建模问题的潜力，在这些应用中，挑战在于收集大量数据集。

创意/价值

在加工参数之间存在复杂关系的情况下，这是首次将灰色模型应用于热喷涂工艺。建模结果表明合理的实验结果和现有的处理知识。