# Electricity Substitution Potential Prediction Based on Tent-CSO-CG-SSA-Improved SVM—A Case Study of China

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

- (1)
- From four perspectives—economic development factors, environmental constraints, technological development and policy influence—we analyzed the influencing factors of electrical power substitution potential and give theoretical guidance for the promotion of electrical power substitution. In the meantime, the potential for electricity substitution is quantified and analyzed quantitatively;
- (2)
- Improvements to the sparrow search algorithm: Firstly, Tent chaotic mapping was used to initialize the sparrow population, so that the sparrow population was diversified. Then, a chicken swarm optimization (CSO) was introduced to improve the global search ability of the algorithm in the individual position update, followed by the introduction of the Cauchy–Gaussian (CG) mutation method to improve the individual fitness, and finally the multi-algorithm improved sparrow search algorithm was obtained;
- (3)
- The multi-algorithm improved sparrow search algorithm was used to optimize the SVM model to forecast the electrical power substitution potential. It provides theoretical support and decision support for China to promote the development strategy of electrical power substitution and achieve sustainable development.

## 2. Factors Influencing and Quantifying the Potential for Electricity Substitution

#### 2.1. Economic Development Factors

#### 2.2. Environmental Constraintss

#### 2.3. Technological Progress Factors

#### 2.4. Policy Influencing Factors

#### 2.5. Electricity Substitution Potential

## 3. Model Design and Algorithm Flow

#### 3.1. Sparrow Search Algorithm

#### 3.2. Multi-Algorithm Improvement of Sparrow Search Algorithm

#### 3.2.1. Tent Chaos Mapping Initialization Population

#### 3.2.2. Chicken Swarm Optimization to Improve Sparrow Search Algorithm

#### 3.2.3. Cauchy–Gaussian Hybrid Mutation

#### 3.3. Support Vector Machine

#### 3.4. Electricity Substitution Potential Prediction Model

- Step 1: Set the parameters related to the sparrow search algorithm with a maximum number of 100 iterations and a population size of 20.
- Step 2: Generate the sparrow population by using Tent chaotic mapping, including the population ratio of both explorers and followers.
- Step 3: Generate new individuals. By introducing the chicken swarm optimization (CSO) to update the follower positions and the Cauchy–Gaussian mutation strategy (CG) to mutate the individuals, the positions of explorers, followers and sparrows under the early warning mechanism are updated. Meanwhile, new solutions are generated continuously and iteratively, and the current optimal solution for the explorer positions and the global optimal solution are recorded.
- Step 4: The fitness function is evaluated to obtain the best adapted individual. SSA should take appropriate measures to ensure the best parameter settings to improve the prediction accuracy of the model. In the SSA, the $K$-fold cross-validation ($K-CV)$ method is used to optimize the model parameters. $K$-fold cross-validation can effectively solve the situation of poor generalization ability of the model overfitting, and the mean square error $MSE$of the prediction model is selected as the fitness function. $K$-fold cross-validation takes advantages of the no-repeat sampling: each individual has only one chance to be included in the training set or test set in each iteration. Combined with the sample size in this paper, $K=5$, where the model evaluation index is ${y}_{i}$, the true value, and ${\widehat{y}}_{i}$ is the predicted value.$$MSE=\frac{SSE}{n}=\frac{1}{n}{{\displaystyle \sum}}_{i=1}^{m}{({y}_{i}-{\widehat{y}}_{i})}^{2}$$
- Step 5: Stop the loop criteria. There are two stopping criteria: one is that the number of iterations reaches the set criteria, and the other is that the model error reaches the expected level. When the iterations meet the stopping criteria, the prediction results are output. Otherwise, repeat the third and fourth steps until the number of iterations reaches the preset value. With cyclic iterations, the best fitness function value is obtained and the best model parameters are obtained. Then, the sparrow search algorithm’s model parameters, $C$and $g$, are updated to the SVM model, and eventually the ISSA-SVM prediction model is obtained.

## 4. Case Study

#### 4.1. Data Sources and Model Evaluation

#### 4.2. Analysis of Prediction Results

## 5. Further Study

## 6. Conclusions

- (1)
- From the perspectives of economic development, environmental constraints, technological progress, and policy support, the forecasting and analysis of electrical power substitution potential in Henan Province were carried out, and quantitative indexes are also given for the above four factors and electrical power substitution potential. The quantitative results show that the electrical power substitution potential of Jiangsu Province and Henan Province in China is increasing year by year, which is consistent with the trend of electrical power substitution strategy promotion.
- (2)
- A forecasting model of electrical power substitution potential is proposed; i.e., the sparrow search algorithm with tent chaos mapping and chicken swarm optimization (CSO) and Cauchy–Gaussian mutation optimization is introduced to optimize the SVM model. For the forecasting of electrical power substitution potential in two Chinese provinces, the data relevant to electrical power substitution potential in Jiangsu Province and Henan Province are used as examples, and both case studies show that the proposed model has improved in prediction accuracy and generalization ability. The mean percentage error (MAPE) of the proposed multi-strategy optimization model is reduced by about five percent compared with that of the single sparrow search algorithm, and the relative error (RE), mean absolute error (MAE), and root mean square error (RMSE) are reduced to different degrees. It can be concluded that the tent chaos mapping, chicken swarm optimization (CSO), and the Cauchy–Gaussian mutation strategy do improve the optimality finding of the sparrow search algorithm, showing that the multi-algorithm hybrid strategy proposed in this paper is successful in predicting the electrical power substitution potential by SVM.
- (3)
- It provides quantitative indicators of the factors influencing the potential for electrical power substitution and forecasts the potential. In addition, it provides theoretical support and a scientific decision-making basis for better guidance of the comprehensive promotion of electrical power substitution, which makes a contribution to the sustainable development of China.

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## Notations and Abbreviations

CSO | Chicken Swarm Optimization |

SSA | Sparrow Search Algorithm |

CG | Cauchy–Gaussian Mutation |

RMSE | Root mean square error |

MAE | Mean absolute error |

MAPE | Mean Absolute Percentage Error |

r | Correlation coefficient |

RE | Relative Error |

SVM | Support vector machines |

Tent | Tent chaos mapping |

ISSA-SVM | Tent-CSO-CG-SSA-SVM |

## References

- Lei, P.; Xiuhui, W.; Zhongfu, T.; Jing, W.; Chengfeng, L.; Weizheng, K. Feasible Electricity Price Calculation and Environmental Benefits Analysis of the Regional Nighttime Wind Power Utilization in Electric Heating in Beijing. J. Clean. Prod.
**2019**, 212, 1434–1445. [Google Scholar] [CrossRef] - Cai, W.; Li, L.; Jia, S.; Conghu, L.; Jun, X.; Luoke, H. Task-Oriented Energy Benchmark of Machining Systems for Energy-Efficient Production. Int. J. Precis. Eng. Manuf. Green Technol.
**2020**, 7, 205–218. [Google Scholar] [CrossRef] - Kumar-Verma, B.; Shirish, S. What Drives Successful Implementation of Pollution Prevention and Cleaner Technology Strategy? The Role of Innovative Capability. J. Environ. Manag.
**2015**, 155, 184–192. [Google Scholar] [CrossRef] - Yanmei, L.; Zeng, C. Study on Regional Electric power Substitution Potential Evaluation Based on TOPSIS Method of Optimized Connection Degree. Power Syst. Technol.
**2019**, 43, 687–695. [Google Scholar] - Ping, G.; Bin, Y. Application of Solid Energy Storage Heating Device in Electric Power Substitution. Power Demand Side Manag.
**2016**, 18, 34–36. [Google Scholar] - Lin, L.; Rong, C.; Lei, F. Investigation of Boilers Energy Alternative Outside the Region of Nanjing Heating Supply. Power Demand Side Manag.
**2010**, 12, 47–49. [Google Scholar] - Ming, L.; Diangang, H.; Youxue, Z. Research and Practice of Renewable Energy Local Consumption Mode in Gansu Province Based on “Double Alternative” Strategy. Power Syst. Technol.
**2016**, 40, 2991–2997. [Google Scholar] - Dongli, C.; Yue, Y.; Zhixiang, L. Application and Efficiency Evaluation of Alternative Energy. J. Mod. Power Syst. Clean Energy
**2011**, 27, 30–34. [Google Scholar] - Feng, Y.; Xiaozhong, X. A Short-Term Load Forecasting Model of Natural Gas Based on Optimized Genetic Algorithm and Improved BP Neural Network. Appl. Energy
**2014**, 134, 102–113. [Google Scholar] [CrossRef] - Nan, L.; Yeyu, Z.; Caiyue, Z. Application of Support Vector Machine Method in Electric Load Forecasting. Power Syst. Technol.
**2007**, 31, 215–218. [Google Scholar] - Yi, L.; Dongxiao, N.; Wei-Chiang, Hong. Short Term Load Forecasting Based on Feature Extraction and Improved General Regression Neural Network Model. Energy
**2018**, 166, 653–663. [Google Scholar] - Ferlito, S.; Adinolfi, G.; Graditi, G. Comparative Analysis of Data-Driven Methods Online and Offline Trained to the Forecasting of Grid-Connected Photovoltaic Plant Production. Appl. Energy
**2017**, 205, 116–129. [Google Scholar] [CrossRef] - Raju, K.; Madurai Elavarasan, R.; Mihet-Popa, L. An Assessment of Onshore and Offshore Wind Energy Potential in India Using Moth Flame Optimization. Energies
**2020**, 13, 3063. [Google Scholar] - Yi, L.; Lee, T. Forecasting Energy Consumption Using a Grey Model Improved by Incorporating Genetic Programming. Energy Convers. Manag.
**2011**, 52, 147–152. [Google Scholar] [CrossRef] - Kostmann, M.; Härdle, W.K. Forecasting in Blockchain-Based Local Energy Markets. Energies
**2019**, 12, 2718. [Google Scholar] [CrossRef] [Green Version] - Yi, S.; Shuang, Z.; Baoguo, S.; Dexiang, J.; Fang, C. Analysis of Electric Power Alternative Potential Under Multi-Scenario. Power Syst. Technol.
**2017**, 41, 118–123. [Google Scholar] - Baoguo, S.; Jia, Z.; Dexiang, J.; Fang, C.; Yi, S. Electric Power Substitution Potential Analysis Method Based on STIRPAT-Ridge regression. Distrib. Util.
**2018**, 35, 68–73. [Google Scholar] - Yi, S.; Mo, S.; Baoguo, S.; Fang, C. Electric Power Substitution Potential Analysis Method Based on Particle Swarm Optimization Support Vector Machine. Power Syst. Technol.
**2017**, 41, 1767–1771. [Google Scholar] - Changzu, L.; Dongxiao, N.; Xinyan, Z.; Bo, M. Zhejiang Province Electric Power Substitution Potential Prediction Model Based on Improved Particle Swarm Optimization BP Neural Network. Int. J. Eng. Sci. Technol.
**2020**, 20, 5173–5179. [Google Scholar] - Wu, J.; Tan, Z.; De, G.; Pu, L.; Wang, K.; Tan, Q.; Ju, L. Multiple Scenarios Forecast of Electric Power Substitution Potential in China: From Perspective of Green and Sustainable Development. Processes
**2019**, 7, 584. [Google Scholar] [CrossRef] [Green Version] - Guojing, L.; Jian, T.; Hu, L.; Xiaoyan, H.; Lei, Z. Forecast of Electric power Substitution Based on Logistic Model. EET
**2018**, 37, 39–43. [Google Scholar] - Yujun, L.; Cheng, H.; Yi, H. Electric power Substitution Potential Prediction Based on Logistic Curve Fitting and Improved BP Neural Network Algorithm. Elektron. Elektrotech.
**2019**, 25, 18–24. [Google Scholar] [CrossRef] - Li, M.-W.; Geng, J.; Wang, S.; Hong, W.-C. Hybrid Chaotic Quantum Bat Algorithm with SVR in Electric Load Forecasting. Energies
**2017**, 10, 2180. [Google Scholar] [CrossRef] [Green Version] - Chenglong, Z.; Shifei, D. A Stochastic Configuration Network Based on Chaotic Sparrow Search Algorithm. Knowl Based Syst.
**2021**, 220, 106924. [Google Scholar] [CrossRef] - Trull, O.; García-Díaz, J.C.; Troncoso, A. One-Day-Ahead Electricity Demand Forecasting in Holidays Using Discrete-Interval Moving Seasonalities. Energy
**2021**, 231, 120966. [Google Scholar] [CrossRef] - Boqiang, L. Structural Changes, Efficiency Improvement and Electricity Demand Forecasting. J. Econ. Res.
**2003**, 05, 57–65. [Google Scholar] - Yancang, L.; Muxuan, H.; Qnglin, G. Modified Whale Optimization Algorithm Based on Tent Chaotic Mapping and Its Application in Structural Optimization. KSCE J. Civ. Eng.
**2020**, 24, 3703–3713. [Google Scholar] - Li, M.; Li, C.; Huang, Z.; Huang, J.; Wang, G.; Liu, P.X. Spiral-Based Chaotic Chicken Swarm Optimization Algorithm for Parameters Identification of Photovoltaic Models. Soft. Comput.
**2021**, 25, 12875–12898. [Google Scholar] [CrossRef] - Yushan, Z.; Guiwu, H. An Analytical Framework for Runtime of a Class of Continuous Evolutionary Algorithms. Comput. Intell. Neurosci.
**2015**, 2015, 485215. [Google Scholar]

Year | Actual Data | ISSA-SVM | SSA-SVM | SVM |
---|---|---|---|---|

2000 | 76.30 | 84.42 | 171.07 | 608.46 |

2001 | 155.44 | 170.54 | 202.75 | 589.98 |

2002 | 246.48 | 230.12 | 246.09 | 584.55 |

2003 | 355.66 | 330.25 | 299.72 | 587.39 |

2004 | 401.35 | 402.99 | 401.76 | 802.31 |

2005 | 409.30 | 412.31 | 441.88 | 662.51 |

2006 | 590.94 | 530.45 | 504.19 | 679.70 |

2007 | 774.91 | 721.36 | 649.17 | 774.52 |

2008 | 807.77 | 840.31 | 835.69 | 927.85 |

2009 | 849.93 | 870.36 | 949.27 | 1023.15 |

2010 | 1185.80 | 1150.41 | 1130.43 | 1186.21 |

2011 | 1414.42 | 1415.42 | 1414.03 | 1402.46 |

2012 | 1582.65 | 1581.32 | 1591.80 | 1545.87 |

2013 | 1921.37 | 1821.31 | 1775.56 | 1661.06 |

2014 | 1908.90 | 1910.63 | 1920.41 | 1733.44 |

2015 | 1957.96 | 2100.14 | 2103.11 | 1801.85 |

2016 | 2215.35 | 2214.25 | 2215.76 | 1794.47 |

2017 | 2523.52 | 2453.41 | 2349.81 | 1796.93 |

2018 | 2840.47 | 2650.31 | 2438.29 | 1800.17 |

2019 | 2883.96 | 2699.89 | 2498.21 | 1752.16 |

ISSA-SVM | SSA-SVM | SVM | |
---|---|---|---|

1 | 0.049 | 0.018 | 23.454 |

2 | 2.857 | 7.392 | 40.435 |

3 | 7.175 | 16.494 | 57.789 |

4 | 6.8178 | 15.441 | 64.594 |

ISSA-SVM | SSA-SVM | SVM | |
---|---|---|---|

RMSE | 61.22 | 130.52 | 391.68 |

r | 0.99 | 0.98 | 0.48 |

MAE | 22.27 | 48.11 | 165.97 |

MAPE | 4.62 | 9.84 | 46.57 |

**Table 4.**Forecast results of electrical power substitution potential in the case of Jiangsu province.

Year | Actual Data | ISSA-SVM | SSA-SVM | SVM |
---|---|---|---|---|

2000 | 0.00 | 31.68 | 27.57 | 128.22 |

2001 | 49.38 | 49.78 | 49.78 | 115.68 |

2002 | 99.53 | 37.85 | 38.35 | 99.93 |

2003 | 80.37 | 38.26 | 39.10 | 82.54 |

2004 | 5.61 | 6.01 | 6.01 | 199.45 |

2005 | 25.98 | 26.38 | 26.38 | 98.93 |

2006 | 51.10 | 79.36 | 88.48 | 97.68 |

2007 | 190.66 | 180.11 | 190.26 | 191.06 |

2008 | 249.48 | 249.88 | 249.88 | 304.73 |

2009 | 290.22 | 399.27 | 416.82 | 430.75 |

2010 | 633.57 | 579.83 | 590.06 | 537.06 |

2011 | 803.75 | 803.35 | 803.35 | 695.61 |

2012 | 850.23 | 822.78 | 817.26 | 778.08 |

2013 | 912.55 | 860.10 | 865.90 | 838.83 |

2014 | 843.61 | 888.52 | 882.87 | 843.21 |

2015 | 853.20 | 857.35 | 853.60 | 829.77 |

2016 | 965.02 | 965.42 | 940.42 | 793.29 |

2017 | 1156.62 | 1157.02 | 1138.22 | 741.08 |

2018 | 1363.55 | 1241.36 | 1183.92 | 757.97 |

2019 | 1342.10 | 1276.86 | 1139.32 | 690.21 |

ISSA-SVM | SSA-SVM | SVM | |
---|---|---|---|

1 | 0.271 | 2.619 | 21.647 |

2 | 0.819 | 1.616 | 56.072 |

3 | 7.135 | 15.172 | 79.894 |

4 | 13.128 | 17.798 | 94.448 |

ISSA-SVM | SSA-SVM | SVM | |
---|---|---|---|

RMSE | 30.97 | 60.96 | 92.23 |

r | 0.97 | 0.91 | 0.74 |

MAE | 9.41 | 21.27 | 92.23 |

MAPE | 3.76 | 9.31 | 60.31 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Geng, J.; Meng, W.; Yang, Q.
Electricity Substitution Potential Prediction Based on Tent-CSO-CG-SSA-Improved SVM—A Case Study of China. *Sustainability* **2022**, *14*, 853.
https://doi.org/10.3390/su14020853

**AMA Style**

Geng J, Meng W, Yang Q.
Electricity Substitution Potential Prediction Based on Tent-CSO-CG-SSA-Improved SVM—A Case Study of China. *Sustainability*. 2022; 14(2):853.
https://doi.org/10.3390/su14020853

**Chicago/Turabian Style**

Geng, Jinqiang, Weigao Meng, and Qiaoran Yang.
2022. "Electricity Substitution Potential Prediction Based on Tent-CSO-CG-SSA-Improved SVM—A Case Study of China" *Sustainability* 14, no. 2: 853.
https://doi.org/10.3390/su14020853