A novel axle temperature forecasting method based on decomposition, reinforcement learning optimization and neural network

Liu, Hui; Yu, Chengming; Yu, Chengqing; Chen, Chao; Wu, Haiping

doi:10.1016/j.aei.2020.101089

Cited by 42 publications

(11 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In an EMD-based hybrid method, a single method (here ARIMA or EWMA) is used for each EMD component which contributes to much computation time. In their hybrid method, [24] found that EWT-Q-BPNN (where EWT stands for Empirical Wavelet Transform, Q stands for Q-learning algorithm and BPNN stands for Back Propagation Neural Network) method required 614.63s while ARIMA required 13.55s for their Series #1 dataset. Here, for BMO dataset, the proposed EMD-ARIMA-EWMA required 1.96s while ARIMA required 0.106s.…”

Section: Discussion and Outcome Of The Studymentioning

confidence: 99%

Improving Stock Price Prediction Using Combining Forecasts Methods

2021

View full text Add to dashboard Cite

This study presents an outcome of pursuing better and effective forecasting methods. The study primarily focused on the effective use of divide-and-conquer strategy with Empirical Mode Decomposition or briefly EMD algorithm. We used two different statistical methods to forecast. One is for the high-frequency EMD components and another for low-frequency EMD components. With two statistical forecasting methods, ARIMA (Autoregressive Integrated Moving Average) and EWMA (Exponentially Weighted Moving Average), we investigated two possible and potential hybrid methods: EMD-ARIMA-EWMA, EMD-EWMA-ARIMA based on high and low-frequency components. We experimented with these methods and compared their empirical results with four other forecasting methods using five stock market daily closing prices from the S&P/TSX 60 Index of Toronto Stock Exchange. This study found better forecasting accuracy from EMD-ARIMA-EWMA than ARIMA, EWMA base methods and EMD-ARIMA as well as EMD-EWMA hybrid methods. Therefore, we believe frequency-based effective method selection in EMD-based hybridization deserves more research investigation for better forecasting accuracy.

show abstract

Section: Discussion and Outcome Of The Studymentioning

confidence: 99%

Improving Stock Price Prediction Using Combining Forecasts Methods

2021

View full text Add to dashboard Cite

show abstract

“…According to the current state S , the agent performs an action a. During this process, the action is selected based on the ԑ -greedy policy as [ 66 ]:

where the parameter

is the exploration probability. Step 4: Calculate the loss function L , get the reward R , and develop the next step strategy.…”

Section: Methodsmentioning

confidence: 99%

“…According to the current state S , the agent performs an action a. During this process, the action is selected based on the ԑ -greedy policy as [ 66 ]:…”

Section: Methodsmentioning

confidence: 99%

A data-driven hybrid ensemble AI model for COVID-19 infection forecast using multiple neural networks and reinforced learning

Jin¹,

Dong²,

Yu³

et al. 2022

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

“…In order to improve the operational safety and efficient planning of the microgrid, it is necessary to forecast the output power of the PV and WT systems (Bochenek et al, 2021;Theocharides et al, 2018). In this work, Machine Learning based method is applied to forecast the meteorological data in a day-ahead, hence, the power output of renewable energy sources will be estimated in order to be fed later to the optimization module (Liu et al, 2020;Nezhad et al, 2020). The estimation of the PV and wind energy is done by the prediction of the temperature, the irradiance and the wind speed which will be used as inputs in the models of PV and WT systems explained in the previous section.…”

Section: Forecasting Modulementioning

confidence: 99%

Energy Management Strategy Based on Marine Predators Algorithm for Grid-Connected Microgrid

Kheiter

Souag

Abdellah

et al. 2022

Int. J. Renew. Energy Dev.

View full text Add to dashboard Cite

This work aims to optimize the economic dispatch problem of a microgrid system in order to cover the load of a commercial building in Algeria. The analyzed microgrid system is connected to the power grid and composed of photovoltaic panels (PV), wind turbine, battery energy storage system (BESS) and diesel generator. To ensure energy balance and the flow of energy, we have implemented an energy management strategy based on Marine Predator Algorithm (MPA) and Multilayer Perceptron Neural Network (MLPNN), which guarantee an optimal economic operation of the system. First, using historical meteorological data, the power generation is forecasted a day-ahead using MLPNN, which allows the optimization of the microgrid operation. Second, the proposed strategy has been studied under three different microgrid configurations. Eventually, the performances of MPA are compared against well-known algorithms. The results indicate that the integration of the PV-BESS microgrid system significantly reduces the daily operating cost up to 34.5%. Due to the availability of wind resources in the studied area, the addition of a wind turbine to the microgrid minimizes the operating cost by 43.96% compared to the operating cost of the power grid. In the case of selling excess energy to the main power grid, the operating cost could be decreased as much as 49.33%.

show abstract

A novel axle temperature forecasting method based on decomposition, reinforcement learning optimization and neural network

Cited by 42 publications

References 43 publications

Improving Stock Price Prediction Using Combining Forecasts Methods

Improving Stock Price Prediction Using Combining Forecasts Methods

A data-driven hybrid ensemble AI model for COVID-19 infection forecast using multiple neural networks and reinforced learning

Energy Management Strategy Based on Marine Predators Algorithm for Grid-Connected Microgrid

Contact Info

Product

Resources

About