Multi-Stage Wiener Process Based Remaining Useful Life Prediction for Load Cells

Brushless DC motors are widely used, and their working temperatures, regarding as degradation processes, are nonlinear and multistage. It is necessary to establish a nonlinear degradation model. In this research, our study was based on accelerated degradation data of motors, which are their working temperatures. A multistage Wiener model was established by using the transition function to modify linear model. The normal weighted average filter (Gauss filter) was used to improve the results of estimation for the model parameters. Then, to maximize likelihood function for parameter estimation, we used numerical optimization method- the simplex method for cycle calculation. Finally, the modeling results show that the degradation mechanism changes during the degradation of the motor with high speed. The effectiveness and rationality of model are verified by comparison of the life distribution with widely used nonlinear Wiener model, as well as a comparison of QQ plots for residual. Finally, predictions for motor life are gained by life distributions in different times calculated by multistage model.

Electronic packaging solder joints are the key parts of mechanical fixation and electrical interconnection between electronic chips and printed circuit boards, which are prone to failure and lead to electronic device failures under the action of vibration environment stress. In regard to the failure characterization and degradation modeling of electronic packaging solder joints under vibration load, this paper adopts environmental stress tests, builds a vibration failure test platform, designs a vibration load excitation spectrum, and obtains solder joint degradation data under vibration stress; it uses the square root amplitude, form factors, and kurtosis factors to characterize the solder joint degradation process, which effectively identify the solder joint degradation node, and improve the data monotonicity of the solder joint degradation process; the Wiener process is used to build a solder joint multi-stage degradation model. Based on the EM algorithm, the hyperparameters of the degradation model have been optimized, and the universal test of the Wiener degradation model with multiple samples is carried out in accordance with the LB index. The analysis shows that the effectiveness of different samples is as high as 87.5%, which verifies the universality of the Wiener degradation model; Based on the crack morphology subject to the solder joint test and the features of the solder joint in the multi-stages, the paper analyzes the crack propagation behavior of the solder joint, and clarifies the failure mechanism of the solder joint.

The rising cost of carrying inventory sets a natural challenge to the widely accepted doctrine that long production runs are efficient. A large variety of process organizations embodied in production/inventory models deal with this problem in the literature. The single-stage lot-size model does not account for the work-in-process inventory, hence it overstates the optimal lot size. Multi-stage manufacturing usually involves work-in-process inventory that increases with the number of production stages. Multi-stage models have been developed for various process organizations. Two models, each representing a class of multi-stage process organization, are examined in this paper. One model uses uniform lot sizes at all stages and allows transportation of equal size portions of the lot between stages. The other model allows a different lot size at each stage, so that only completed lots are transported to the next stage. The costs of the two representative models are compared for a large number of simul...

In multi-stage processes, dependence on noisy observations of the intermediates is a problem to overcome to predict the outputs accurately. This requires a Multi-Stage Gaussian Process (MSGP)- a modeling idea to incorporate such intermediate observations, considering various observation likelihoods effectively. The MSGP may further boost predictive performance by indirectly observing the multi-stage process by adopting Directed Acyclic Graph (DAG) architecture and Variational Inference (VI) methods. Such a model would use the prior information and increase the accuracy of inference, making Bayesian optimization and prediction effective in situations where one can hardly make direct observations.

Effective referral triage in pain medicine is essential to ensure that patients receive timely and appropriate care. This study presents a multistage machine learning framework to better identify patients who may benefit from one of five specialised pain procedures. Two years of clinic data were used for training and validation, consisting of 231 features for 3,552 patients. The proposed approach started with the baseline model selection, followed by stage iterations acquiring the elbow method for stage decision. The Easy Ensemble was selected among methods and applied for each stage. Results showed that the approach improves prediction accuracy in the true positive rate (TPR) and area under the curve (AUC). The final-stage models achieved the improvement over single-stage model by as much as 34.8% and 23% on TPR and AUC. This multistage framework can enhance triage accuracy and hold potential for broader application in other clinical settings.

Degradation models are widely used to assess the lifetime information of highly reliable products possessing quality characteristics that both degrade over time and can be related to reliability. The performance of a degradation model largely depends on an appropriate model description of the product's degradation path. Conventionally, the random or mixed-effect model is one of the most well-known approaches presented in the literature in which the normal distribution is commonly adopted to represent unit-to-unit variability in the degradation model. However, this assumption may not appropriately signify accurate projections for practical applications. This paper is motivated by laser data wherein the normal distribution is relaxed with a skew-normal distribution that consequently provides greater flexibility as it can capture a broad range, non-normal, asymmetric behavior in unit-to-unit variability. Based on the proposed degradation model, we first derive analytical expressions for a product's lifetime distribution along with its corresponding mean-time-to-failure ( <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MTTF</i> ). We then utilize the laser data to illustrate advantages gained by the proposed model. Finally, we address effects from the skewness parameter with regard to the accuracy of both a product's <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MTTF</i> and its <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">q</i> th failure quantile; especially when the underlying skew-normal distribution is mis-specified as a normal distribution. The result demonstrates that effects from the skewness parameter on the tail probabilities of a product's lifetime distribution are not negligible when the random effect of the true degradation model follows a skew-normal distribution.

In this work a one dimensional steady state model is developed for a single stage and two-stage bottom fed entrained flow coal gasifier for. The single stage model was first analysed for two different oxidants (i) oxygen and (ii) air to study their effects on gasification. Analysis proved oxygen to be the better oxidant. The model is then extended to a two stage gasifier. Here 30 (70) percent of the coal is fed in the first stage (second stage). The first (second) stage operates in oxidant rich (lean) environment. The performances of single stage and two-stage models are compared in terms of their cold gas efficiencies for the same coal feed rates. It is observed that a two-stage system has better cold gas efficiency (77%) than a single stage system (72%) with lower oxygen consumption. The two-stage model is used to optimise the O2/Coal ratio as the H2O/Coal ratio is varied in the 2nd stage. The optimum yields the highest cold gas efficiency with minimum possible oxygen consumption.

An aero engine fan rotor is composed of a multi-stage disk and multi-stage blades. Excessive unbalance of the aero engine fan rotor after assembly is the main cause of aero engine vibration. In the rotor assembly process, blade sequencing optimization and multi-stage blade set assembly phase optimization are important for reducing the overall rotor unbalance. To address this problem, this paper proposes a semi-physical simulation method based on reinforcement learning to optimize the balance in the fan rotor assembly process. Firstly, based on the mass moments of individual blades, the diagonal mass moment difference is introduced as a constraint to build a single-stage blade sorting optimization model, and reinforcement learning is used to find the optimal sorting path so that the balance of the single-stage blade after sorting is optimal. Then, on the basis of the initial unbalance of the disk and the unbalance of the single-stage blade set, a multi-stage blade assembly phase optimization model is established, and reinforcement learning is used to find the optimal assembly phase so that the overall balance of the rotor is optimal. Finally, based on the collection of data during the assembly of the rotor, the least-squares method is used to fit and calculate the real-time assembly unbalance to achieve a semi-physical simulation of the optimization of balance during the assembly process. The feasibility and effectiveness of the proposed method are verified by experiments.

The aim of this paper is to price an American option in a fuzzy binomial tree model of the investment evaluation of land expropriation, to offer a theoretical support for reasonable determination on land expropriation. American option valuation is usually performed, under the risk-neutral valuation paradigm, by using numerical procedures such as the binomial option pricing model of Cox et al. (1979). A key input of the fuzzy binomial model is the volatility of the underlying asset, which is an unobservable parameter. Thus, in this paper we use a possibility distribution in order to model the uncertainty on the volatility. Structure of the paper is as follows: firstly, the paper builds the one-period binary tree model, analyze the mean and variance of trapezoidal fuzzy numbers; secondly, the paper builds single-stage fuzzy binomial model and multi-stage fuzzy binomial model; finally, a numerical example is studied.

Acquiring sonic waves is an essential part of oil and gas exploration as they give critical information about the well’s data and lithography at each well depth progression. However, these measurements are not always accessible, making analysis challenging. As computational power has improved, machine learning methods may now be used to predict these values from other data. Nonetheless, one shortcoming of existing models is that the most of them are not transparent (i.e., black box models). As a result, although promising great performance, they do not offer much insights to petrophysicists and geologists. This research aims to generate mathematical models for predicting compression wave (P-wave) and shear wave (S-wave) readings using a multistage evolutionary modelling approach. In particular, a multistage equation modelling approach using tree-based Genetic Programming (GP) and Adaptive Differential Evolution (ADE) is proposed. The obtained best mathematical models yield R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.745 and 0.9066 for P-wave and S-wave regression on normalized data, respectively. The average performance of models are R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.90 (P-Wave) and R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.75 (S-Wave). The performance of these mathematical models are comparable with other ”black box” models but with more compact mathematical approach in regression thereby opening opportunities for interpretability and analysis. Finally, the ”white box” models presented in this paper can be fine-tuned further as needed.

As the core component responsible for high-frequency power switching in photovoltaic inverters, accurately predicting the remaining useful life (RUL) of insulated gate bipolar transistors (IGBTs) has become a key factor in ensuring the stable operation of photovoltaic systems. However, existing methods struggle to precisely characterize the degradation characteristics and processes of IGBTs at different time points. To address these issues, this paper proposes a MIG-PI-Pathformer RUL prediction method that integrates physical degradation models with deep learning. This method establishes a multi-stage Inverse Gaussian degradation model based on the physical failure mechanisms of IGBTs and couples it with the dual attention mechanism of the Pathformer model to capture complex degradation features, adaptively divide time scales, and thereby correct prediction errors in the physical model; Additionally, physical rule constraints are incorporated into the Pathformer loss function to ensure that RUL predictions align with degradation mechanisms. Simulation results show that, on NASA’s IGBT aging dataset, compared to the single Pathformer, the proposed method reduces MSE and MAE by 70.21% and 17.84%, respectively, and improves R2 by 7.66%. This method provides more accurate and physically interpretable technical support for fault warning and optimized maintenance of photovoltaic inverters.

This paper considers the Wiener degradation model with random effects. Random-effect models take into account the unit-to-unit variability of the degradation index. It is assumed that a random parameter has a truncated normal distribution. During the research, the expression for the maximum likelihood estimates and the reliability function has been obtained. Two statistical tests have been proposed to reveal the existence of random effects in degradation data corresponding to the Wiener degradation model. The first test is a well-known likelihood ratio test, and the second one is based on the variance estimate of the random parameter. These tests have been compared in terms of power with the Monte-Carlo simulation method. The result of the research has shown that the criterion based on the variance estimate of the random parameter is more powerful than the likelihood ratio test in the case of the considered pairs of competing hypotheses. An example of the analysis using the proposed tests for the turbofan engine degradation data has been considered. The data set includes the measurements recorded from 18 sensors for 100 engines. Before constructing the degradation model, the single degradation index has been obtained using the principal component method. The hypothesis of the random effect insignificance in the model has been rejected for both tests. It has been shown that the random-effect Wiener degradation model describes the failure time distribution more accurately than the fixed-effect Wiener degradation model.

A measurement system of contact force between overhead contact line and pantograph of train is developed which measures the contact force by using four sets of full-bridge strain gauges instead of load cells and accelerometers. The sensors are installed on the pan head of pantograph and the measured data from the sensors are transmitted to a server system in the train by way of wireless Lan. This configuration of the measuring system makes it easy to install on the trains without any alteration of train system. The measurement system is applied to KTX on the Kyungbu high speed line, and the measured contact force data shows good agreement with those measured by load cell and accelerometers. The waveform of the contact force between overhead contact line and pantograph contains essential information about their conditions. The proposed measurement system can probe any defects on overhead contact lines with train running at high speed, which will be a powerful solution for the maintenance of long-distance overhead contact lines.

As improved efficiency goals require ever smaller gaps between rotating turbomachinery blade tips and stationary outer casings, it becomes more important to understand the physics of what happens when a blade tip rubs against the casing. A key piece of information to be determined is the force exerted on the blade tip during the rub event. This paper presents a method of extracting the blade tip rub forces from experiments conducted at The Ohio State University Gas Turbine Laboratory. To identify the forces during the rub, three multi-axis load cells are mounted behind a segment of the casing to measure the forces exerted on the casing during the rub event. Recovering blade tip forcing from this load cell data requires creating an appropriate model that can relate the measured data at the load cells to the highspeed rotating blade loads. The continuous, repeated nature of the rubs in the experiment makes this a challenge. This paper discusses the creation and performance of a state-space realization model developed from stationary ping test data that is used to address these challenges. The model is augmented with information on the angular position of the blades from a rotary encoder and information on the angular span of the rub, which increases with rub depth. Challenges and solutions related to finding tangential frequency responses, simultaneous blade rubs, and irregularities in the experimental data are also discussed. This paper aims to summarize the process and effectiveness of this approach for finding blade tip forces, as well as lessons learned along the way.

GMF-GAN: Gradual multi-granularity semantic fusion GAN for text-to-image synthesis