BP neural network prediction for Si and S contents in hot metal of COREX process based on mathematical analysis and Deng’s correlation

J of Ultrasound Medicine

Meng

et al. 2023

ObjectiveThe purpose of this study is to accurately monitor temperature during microwave hyperthermia. We propose a temperature estimation model BP‐Nakagami based on neural network for Nakagami distribution.MethodsIn this work, we designed the microwave hyperthermia experiment of fresh ex vivo pork tissue and phantom, collected ultrasonic backscatter data at different temperatures, modeled these data using Nakagami distribution, and calculated Nakagami distribution parameter m. A neural network model was built to train the relationship between Nakagami distribution parameter m and temperature, and a BP‐Nakagami temperature model with good fitting was obtained. The temperature model is used to draw the two‐dimensional temperature distribution map of biological tissues in microwave hyperthermia. Finally, the temperature estimated by the model is compared with the temperature measured by thermocouples.ResultsThe error between the temperature estimated by the temperature model and the temperature measured by the thermocouple is within 1°C in the range of 25°C–50°C for ex vivo pork tissue, and the error between the temperature estimated by the temperature model and the temperature measured by the thermocouple is within 0.5°C in the range of 25°C–50°C for phantom.ConclusionsThe results show that the temperature estimation model proposed by us is an effective model for monitoring the internal temperature change of biological tissues.

Section: Backscattered Echo Assessmentmentioning

confidence: 99%

Temperature Monitoring During Microwave Hyperthermia Based on BP‐Nakagami Distribution

J of Ultrasound Medicine

Meng

et al. 2023

Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023)

“…Also in 2020, Xin Zhang proposed a BP neural network model for financial market trend [7]. In 2021, Zhou Heng along with his team proposed a prediction model for Si and S contents in hot metal of COREX process based on mathematical analysis and Deng's correlation [5]. In 2022, Sun Gang and his team applied BP neural network in prediction model of Starch-based aerogel prepared by freeze-drying [4].…”

Section: Introductionmentioning

confidence: 99%

A word's distribution of reported results prediction model based on BP neural network

Yin,

Ning,

Huang

2023

In recent years, COVID-19 has brought a huge impact on people's life, making online activities a new trend, which promotes the rapid development of mini-games. A mini-game called "Wordle" came out currently has been brought into focus. Players are required to solve the puzzle by guessing a five-letter word in six tries or less, receiving feedback with every guess. In this paper, we use BP neural network to predict the distribution of reported results for a given solution word on a certain future date(March 1, 2023 as an example) by using the daily data from January 7, 2022 to December 31, 2022. The empirical result shows that the distribution of reported results will be (1: 10%, 2: 15%, 3: 35%, 4: 25%, 5: 10%, 6: 5%, X: 0%). Besides, the R-values of the four sets of regression data are 0.92895, 0.904, 0.89493 and 0.9199 respectively, which shows that it is effective to use BP neural network to predict the distribution of reported results.

Prediction model of BOF end-point temperature and carbon content based on PCA-GA-BP neural network

Cheng

2022

Metall. Res. Technol.

Accurate prediction of temperature and carbon content of liquid steel plays an important role in steelmaking process. In order to enhance the accuracy of predicting the basic oxygen furnace (BOF) end-point temperature and carbon content of liquid steel, a hybrid model based on principal component analysis (PCA) − genetic algorithm (GA) − backpropagation (BP) neural network is proposed. PCA is used to reduce the dimensionality of the input variables and eliminate the collinearity among the variables, then the obtained principal components are seen as new input variables of the BP neural network. GA is employed to optimize the initialized weights and thresholds of the BP neural network. Data from a 250t BOF of H steel plant in China is used to test and validate the model. The results show that the prediction accuracy of the single output models is higher than that of the dual output models. The PCA-GA-BP neural network model with single output shows higher prediction performance than others. The root mean square error of temperature between predicted and actual values is 7.89, and that of carbon content is 0.0030. Therefore, the model can provide a good reference for BOF end-point control.