Machine learning-based prediction of component self-alignment in vapour phase and infrared soldering

Abstract

Purpose This paper aims to investigate the self-alignment of 0603 size (1.5 × 0.75 mm) chip resistors, which were soldered by infrared or vapour phase soldering. The results were used for establishing an artificial neural network for predicting the component movement during the soldering. Design/methodology/approach The components were soldered onto an FR4 testboard, which was designed to facilitate the measuring of the position of the components both prior to and after the soldering. A semi-automatic placement machine misplaced the components intentionally, and the self-alignment ability was determined for soldering techniques of both infrared and vapour phase soldering. An artificial neural network-based prediction method was established, which is able to predict the position of chip resistors after soldering as a function of component misplacement prior to soldering. Findings The results showed that the component can self-align from farer distances by using vapour phase method, even from relative misplacement of 50 per cent parallel to the shorter side of the component. Components can self-align from a relative misplacement only of 30 per cent by using infrared soldering method. The established artificial neural network can predict the component self-alignment with an approximately 10-20 per cent mean absolute error. Originality/value It was proven that the vapour phase soldering method is more stable from the component’s self-alignment point of view. Furthermore, machine learning-based predictors can be applied in the field of reflow soldering technology, and artificial neural networks can predict the component self-alignment with an appropriately low error.

Parameter optimization for surface mounter using a self-alignment prediction model

“…After comparing the loss function at different folds, the number of folds was chosen. GridSearchCV exhaustively considers all parameter combinations and selects the model whose parameter combination gives the best result (Krammer et al, 2019).…”

“…As a result, using small-size passive chip components in electronic packaging became necessary, especially in consumer electronic products (Najib et al , 2018). SMT made it possible to fit several small electronic components such as capacitors and resistors on a compact PCB (Gengenbach et al , 2020; Illés et al , 2020).…”

A pick-and-place process control based on the bootstrapping method for quality enhancement in surface mount technology

“…Due to the complexity and variability of manufacturing processes, data-driven methods are employed to study self-alignment. In [20], an artificial neural network (ANN) was utilized to predict the self-alignment of 0603-size chip resistors soldered using infrared or vapor phase soldering. Components were intentionally displaced from 0 to 800 µm in width and 0 to 300 µm in length.…”

Deep Forest Regression for Short-Term Load Forecasting of Power Systems

“…Except for ANNs and DNNs, numerous algorithms have been applied to predicted models. For example: Levenberg-Marquardt has been employed to estimate the state-of-charge of lithium-ion batteries [23]; Conjugate gradient with Powell/Beale restarts have been applied to plan the path of catering robots [24]; Polak-Ribiere conjugate gradient has been utilized to accumulate the global convergence of nonconvex functions [25]; Fletcher-Powell conjugate gradient has been used to predict component self-alignment [26]; One step secant has been applied to train a cascade ANN [27]; Resilient Backpropagation has been employed to improve the optical coherent transmission [28]; Bayesian regularization has been utilized to solve the global optimization problems [29]; Variable learning rate gradient descent has been employed to regulate the weight and threshold values of layers [30]; Support vector machine regression (Gaussian) has been utilized for heating and cooling load predictions [31]; Linear programming boosting has been employed to non-intrusive load monitoring systems [32]; Adaptive boosting has been improved to automatic wireless signal classification [33]; Extra trees classifier has been applied into natural language processing [34]; Broyden-Fletcber-Goldfarb-Shanno Quasi-Newton has been applied for brain image segmentation [35]; Moving average method has been employed to predict the solar power outputs [36]; Decision tree has been utilized to predict high-risk kidney transplantation [37]; Random subspace binary and multi-class has been applied for disease diagnosis [38]; Support vector machine regression (Linear) has been utilized to predict multi-parameter manufacturing quality [39]; Multiple proportion smoothing method has been applied to the STLF [40]; Random under sampling boosting has been employed to detect non-technical losses of electric distribution systems [41].…”