Low-Cost Pyrometry System With Nonlinear Multisense Partial Least Squares

Accurate high-temperature measurement is very important for process monitoring of an industrial system. Infrared thermometers usually can handle no more than 1000 °C and should use some expensive accessories for higher temperature measurements. This paper proposes a low-cost pyrometry system with nonlinear multisense partial least squares (NMSPLS). The ordinary camera with different filters is designed to collect the images of hot object at different wavelengths, and the NMSPLS is presented for predicting the temperature of the hot object from the obtained images. For the proposed method, the obtained images are represented by the multisense tensor, where red, green, and blue are regarded as three different dimensions in a sense of the tensor, respectively. The proposed method integrates an outer model and a nonlinear inner model. For the outer model, the independent variables and the dependent variables are projected into a low-dimensional common latent subspace. The weight matrices are calculated from the independent variables by the tucker decomposition, and the single value decomposition is adopted for extracting the latent variables (Lvs) based on the covariance between the independent variables and the dependent variables. For the nonlinear inner model, the neural network is adopted and the extracted Lvs are used as the input and the output of the neural network, respectively. Two real experiments are performed for estimating the proposed method. The experimental results verify that the proposed method can be applied for pyrometry and have higher effectiveness.

I. Introduction

Temperature is an important parameter of industrial process, and may reach several thousand centigrade, for example, the temperature of welding molten pool [1]. Therefore, high temperature limits the application of the traditional sensors for the contact temperature measurement in the field [2]. Based on the Planck’s radiation law, the radiation intensity of a hot body at any wavelength has the nonlinear relationship with the temperature and the emissivity of the hot body [3], [4]. The least-squares-based multiwavelength pyrometry model is presented for the noncontact temperature measurement [5], [6]. However, due to the change of the emissivity, the model between the temperature and the radiations intensity may be difficult to be built by the derivation of the formula. Furthermore, there was not a generalization model for different gray bodies in field [7]. Because the imaging technology could be adopted for the noncontact temperature measurement, the ratio pyrometry with the emissivity being a constant is proposed. Infrared thermometer is a popular temperature sensor based on the ratio pyrometry, and can usually handle no more than 1000 centigrade [8]. For high-temperature measurement, the infrared thermometer need to be equipped some expensive accessories, for example, the cooling device, in order to prevent the pixel from being burnt [9], [10]. Because the images of hot object in the visible range could represent the temperature under a certain extent [11], building the high-temperature prediction model based on the regression analysis for the images in the visible/near infrared range could realize the low-cost pyrometry system.