【摘要】 为适应吉林省未来电力发展的“双碳”转型目标，考虑新形势下多元影响因素的作用，需要融合社会化与碳排放相关影响因素，以提高电力需求预测的精确度。在此背景下，现有模型在电力需求预测的稳定性和准确度方面面临挑战。为应对这些挑战，首先通过系统动力学模型对影响电力需求的多个因素进行分析，基于相关性分析，进一步筛选对电力需求产生显著影响的关键指标，确定常住人口、工业增加值、能源消费总量、能源消费结构低碳化指数、人均GDP和GDP六个强关联指标。增加对碳排放指标的引入，突显在“双碳”方面的创新性关注；随后，采用粒子群优化算法(Particle Swarm Optimization, PSO)对支持向量机(Support Vector Machines, SVM)模型的关键参数进行优化，并构建PSO-SVM电力需求预测模型，从而克服现有模型容易陷入局部最优解的问题。通过算例分析分别与传统SVM、BP模型和优化过的PSO-BP模型对比，验证PSO-SVM预测模型的有效性。在电力预测方面，该模型不仅具有较高的准确度，还呈现出更快的训练速度；最终，将该预测模型应用于吉林省2023—2028年电力需求预测，为电力规划与决策提供有力的支持与参考。更多还原

【Abstract】 In order to adapt to the “dual carbon” transformation goal of future electric power development in Jilin Province, considering the role of multiple influencing factors under the new situation, it is necessary to integrate social and carbon emission related influencing factors to improve the accuracy of power demand forecasting.In the current context, the existing models are still facing challenges in terms of stability and accuracy of electricity demand forecasting.In order to address these challenges, firstly, multiple factors affecting power demand are analyzed through system dynamics model.Based on rigorous correlation analysis, key indicators that have a significant impact on power demand are further screened.Six strongly related indicators, namely permanent population, industrial added value, total energy consumption, low-carbon index of energy consumption structure, per capita GDP and GDP,were determined, and the introduction of carbon emission indicators was increased, highlighting the innovative attention in the “double carbon” aspect.Then, Particle Swarm Optimization(PSO) was used to optimize the key parameters of the Support Vector Machines(SVM) model, and the PSO-SVM power demand prediction model was constructed.The problem that the existing model is easy to fall into the local optimal solution is overcome.The effectiveness of the PSO-SVM model is verified by comparison with the traditional SVM model, BP model and the optimized PSO-BP model.In power forecasting, the model not only has high accuracy, but also shows a faster training speed.Finally, the forecast model is applied to the power demand forecast of Jilin Province from 2023 to 2028,which provides a strong support and reference for power planning and decision-making.更多还原