Scholarlink

<p>The scenario evolution of forest fires is highly complex. Scenario deduction analysis can analyze the evolution mechanism of forest fires from a microscopic perspective, thereby deducing the possible scenarios that may occur in the future. This serves as an important decision-making basis for decision-makers to effectively respond to forest fires. The accuracy of forest fire scenario deduction directly affects the effectiveness of subsequent decision-making and response. Therefore, this paper proposes a forest fire scenario deduction method based on the Belief Rule Base (BRB) to improve the accuracy of scenario deduction results. Firstly, a forest fire scenario deduction model is constructed, which effectively helps to understand the scenario evolution mechanism of forest fires from a microscopic perspective. Then, the scenario deduction model is deeply integrated with the BRB to establish a forest fire scenario deduction model based on the BRB. To ensure the accuracy of the deduction results of this model, the Spearman rank correlation coefficient is used in the model to identify the attributes of scenario elements, and the Cartesian product of attribute reference values and expert knowledge is utilized to construct an effective belief rule base. Subsequently, the information gain and differential evolution methods are employed to objectively optimize the model parameters. Finally, taking the Fairview forest fire incident in Southern California, the United States in 2022 as a case, the accuracy and effectiveness of the proposed method in this paper are verified. The experimental results show that the scenario deduction results obtained by proposed method are close to the actual results, demonstrating good accuracy performance. To further verify the generalization performance of the method in this paper, 347 pieces of forest fire data are subjected to five-fold cross-validation. Compared with nine mainstream scenario deduction methods, it shows certain advantages in terms of deduction accuracy. The method in this paper can more accurately deduce forest fire scenarios, thus providing a reference for decision-makers to understand the development trend of forest fires, and further providing decision support for more effectively formulating response strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A novel forest fire scenario deduction method based on belief rule base

  • 期刊论文

  • 发表日期：

    2026年7月21日

查看全文
  • Rui Zhao,
  • Ning Wang,
  • Wenguang Yan,
  • Enhui Zhao

摘要

The scenario evolution of forest fires is highly complex. Scenario deduction analysis can analyze the evolution mechanism of forest fires from a microscopic perspective, thereby deducing the possible scenarios that may occur in the future. This serves as an important decision-making basis for decision-makers to effectively respond to forest fires. The accuracy of forest fire scenario deduction directly affects the effectiveness of subsequent decision-making and response. Therefore, this paper proposes a forest fire scenario deduction method based on the Belief Rule Base (BRB) to improve the accuracy of scenario deduction results. Firstly, a forest fire scenario deduction model is constructed, which effectively helps to understand the scenario evolution mechanism of forest fires from a microscopic perspective. Then, the scenario deduction model is deeply integrated with the BRB to establish a forest fire scenario deduction model based on the BRB. To ensure the accuracy of the deduction results of this model, the Spearman rank correlation coefficient is used in the model to identify the attributes of scenario elements, and the Cartesian product of attribute reference values and expert knowledge is utilized to construct an effective belief rule base. Subsequently, the information gain and differential evolution methods are employed to objectively optimize the model parameters. Finally, taking the Fairview forest fire incident in Southern California, the United States in 2022 as a case, the accuracy and effectiveness of the proposed method in this paper are verified. The experimental results show that the scenario deduction results obtained by proposed method are close to the actual results, demonstrating good accuracy performance. To further verify the generalization performance of the method in this paper, 347 pieces of forest fire data are subjected to five-fold cross-validation. Compared with nine mainstream scenario deduction methods, it shows certain advantages in terms of deduction accuracy. The method in this paper can more accurately deduce forest fire scenarios, thus providing a reference for decision-makers to understand the development trend of forest fires, and further providing decision support for more effectively formulating response strategies.