A hesitant fuzzy wind speed forecasting system with novel defuzzification method and multi-objective optimization algorithm
Introduction
Industrialization consumes a large amount of energy reserves, and this has caused countries worldwide to face the current energy shortage situation. The massive consumption of nonrenewable resources not only results in polluting gases being emitted in the surrounding air, but it also causes serious damage to the ecology and vegetation. Wind power can help alleviate the pressure of the energy crisis and environmental pollution. The share of wind energy in the energy supply of various communities is increasing, and it can provide energy in many modern cities and societies. Accurately forecasting wind speed can help to find appropriate solutions to the optimal operation and planning problems of different parts of the power system (such as the distribution and transmission parts) and to optimize the design of wind farms (Dong et al., 2017, Pearre and Swan, 2018). At the same time, it can significantly reduce the risk of using wind energy in the power system and increase the penetration rate of the renewable energy resource (RES).
The biggest difference between wind energy and other energy sources is the fuzziness, randomness, and uncontrollability of wind speed (Li et al., 2019). Wind power is generated from the wind, which has the characteristics of volatility and intermittency, which makes the dispatch and control of wind farms more difficult.
The objective of this study was to develop an accurate and effective wind speed forecasting system that can better deal with the fuzziness, randomness, and nonlinearity of wind speed series. Owing to the advantages of machine-learning models and fuzzy-logic models, a hesitant fuzzy wind speed forecasting system with a novel aggregate information method and multiobjective optimization algorithm is proposed. It mainly consists of three modules: a data preprocessing module, a fuzzy computation and forecasting module, and a results evaluation module.
In the data preprocessing module, an improved empirical mode decomposition (EMD) powered by the soft sifting stopping criterion is applied to the preprocessing of the wind speed series. It has strong robustness. In the fuzzy computation and forecasting module, the rate of change of the wind speed series is calculated to reduce the impact of data magnitude on forecasting accuracy. Multifuzzification methods (an equal-interval-based method and equal-frequency-based method) are proposed to fuzzify the datasets. This can solve the problem of the nondeterministic nature of fuzzy time series (FTS). Moreover, a multiobjective optimization algorithm is employed to determine the optimal weights of different intervals accurately and stably. Furthermore, a defuzzification approach that integrates a combined midpoint with an ordered weighted averaging (OWA) operator and a regular increasing monotone (RIM) quantifier are proposed to calculate the final forecasting results. In the results evaluation module, the out-of-sample R2 is computed to compare the performance of the proposed system with the comparison models, and the Diebold–Mariano (DM) test is carried out to highlight the difference in accuracy between the proposed system and the benchmark models.
The remainder of this article is organized as follows. Section 2 provides a review of prior work on wind speed forecasting. Section 3 presents the details of the proposed wind speed forecasting system with a novel aggregate information method and multiobjective optimization algorithm. Section 4 introduces the experimental preparation. In Section 5, the experimental results on wind speed forecasting datasets are analyzed. Further comparisons are discussed in Section 6. Finally, the key conclusions are summarized in Section 7.
Section snippets
Literature review
To obtain accurate forecasting results, a variety of approaches have been developed. Table 1 shows the current wind speed forecasting methods. It can be summarized as follows. (1) Physical forecasting approaches have an advantage in long-term wind speed forecasting, but they require a large number of data, and information should be described in detail based on the conditions of the wind farm (Chen et al., 2018). (2) Statistical models are easy to model and lack the ability to process nonlinear
Design of the wind speed forecasting system
A novel wind speed forecasting system is proposed that contains three main modules: a data preprocessing module, a fuzzy computation and forecasting module, and a results evaluation module. The framework of the proposed wind speed forecasting system is shown in Fig. 1.
Experiment preparation
In this section, detailed information about the experiments is introduced, including a description of the datasets, evaluation criteria, and the experimental arrangement.
Wind speed forecasting with different time resolutions
The wind speed series was first decomposed into several components, and the highest-frequency component was regarded as a random noise impact. Then, to better extract the features of the wind speed series and smooth the original series, the high-frequency component was eliminated, and the series was reconstructed to carry out experiments. Fig. 2 depicts the decomposition and feature selection process of the 10-min dataset at site 1. The figure shows that the reconstructed series generated by
Further discussion
In order to further verify the effectiveness and efficiency of the proposed system, this section discusses the statistical test and the operation time of the system.
Conclusion
The uncertainty, nonlinearity, and fuzzy characteristics embedded in wind speed series lead to unsatisfactory results, and few studies have been conducted on aggregating hesitant fuzzy information during wind speed forecasting. This study improved the conventional FTS forecasting models based on HFSs and developed a wind speed forecasting system based on hesitant fuzzy information for the first time. Multifuzzification methods were proposed to fuzzify the universe of discourse into equal and
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This project is supported by National Natural Science Foundation of China (Grant No. 71671029) and High-level Innovation Team Overseas Training Project of Department of Education of Liaoning province in China (No. 2018LNGXGJWPY-ZD005).
References (63)
- et al.
Repeated wavelet transform based ARIMA model for very short-term wind speed forecasting
Renewable Energy
(2019) - et al.
Fuzzy time series forecasting method based on hesitant fuzzy sets
Expert Systems with Applications
(2016) - et al.
A new fuzzy time series forecasting model combined with ant colony optimization and auto-regression
Knowledge-Based Systems
(2015) - et al.
Wind speed forecasting using nonlinear-learning ensemble of deep learning time series prediction and extremal optimization
Energy Conversion and Management
(2018) - et al.
Short-term wind speed prediction using an unscented Kalman filter based state-space support vector regression approach
Applied Energy
(2014) Forecasting enrollments based on fuzzy time series
Fuzzy Sets and Systems
(1996)- et al.
Short-term wind forecast of a data assimilation/weather forecasting system with wind turbine anemometer measurement assimilation
Renewable Energy
(2017) - et al.
A new combined model based on multi-objective salp swarm optimization for wind speed forecasting
Applied Soft Computing
(2020) - et al.
Point and interval forecasting for wind speed based on linear component extraction
Applied Soft Computing
(2020) - et al.
A novel forecasting model based on a hybrid processing strategy and an optimized local linear fuzzy neural network to make wind power forecasting: A case study of wind farms in China
Renewable Energy
(2017)
A novel hybrid model for short-term wind power forecasting
Applied Soft Computing
ARMA based approaches for forecasting the tuple of wind speed and direction
Applied Energy
Wind speed prediction using the hybrid model of wavelet decomposition and artificial bee colony algorithm-based relevance vector machine
International Journal of Electrical Power & Energy Systems
A novel wind speed forecasting model based on moving window and multi-objective particle swarm optimization algorithm
Applied Mathematical Modelling
Research and application of a combined model based on frequent pattern growth algorithm and multi-objective optimization for solar radiation forecasting
Applied Energy
Effective lengths of intervals to improve forecasting in fuzzy time series
Fuzzy Sets and Systems
Short-term wind speed forecasting framework based on stacked denoising auto-encoders with rough ANN
Sustainable Energy Technologies and Assessments
Variable weights combined model based on multi-objective optimization for short-term wind speed forecasting
Applied Soft Computing
A hybrid forecasting system based on fuzzy time series and multi-objective optimization for wind speed forecasting
Applied Energy
An innovative hybrid system for wind speed forecasting based on fuzzy preprocessing scheme and multi-objective optimization
Energy
Research and application of a combined model based on variable weight for short term wind speed forecasting
Renewable Energy
A novel two-stage deep learning wind speed forecasting method with adaptive multiple error corrections and bivariate Dirichlet process mixture model
Energy Conversion and Management
Short-term wind speed forecasting based on the Jaya-SVM model
International Journal of Electrical Power & Energy Systems
Orness and parameterized RIM quantifier aggregation with OWA operators: A summary
International Journal of Approximate Reasoning
Probabilistic spatiotemporal wind speed forecasting based on a variational Bayesian deep learning model
Applied Energy
A combined forecasting model for time series: Application to short-term wind speed forecasting
Applied Energy
Time-frequency representation based on robust local mean decomposition for multicomponent AM-FM signal analysis
Mechanical Systems and Signal Processing
Salp Swarm Algorithm: A bio-inspired optimizer for engineering design problems
Advances in Engineering Software
Statistical approach for improved wind speed forecasting for wind power production
Sustainable Energy Technologies and Assessments
Wind speed prediction method based on Empirical Wavelet Transform and New Cell Update Long Short-Term Memory network
Energy Conversion and Management
A novel deep learning ensemble model with data denoising for short-term wind speed forecasting
Energy Conversion and Management
Cited by (36)
Novel wind-speed prediction system based on dimensionality reduction and nonlinear weighting strategy for point-interval prediction
2024, Expert Systems with ApplicationsMulti-rule combination prediction of compositional data time series based on multivariate fuzzy time series model and its application
2024, Expert Systems with ApplicationsHybrid model for profit-driven churn prediction based on cost minimization and return maximization
2023, Expert Systems with ApplicationsBetter wind forecasting using Evolutionary Neural Architecture search driven Green Deep Learning
2023, Expert Systems with Applications