改进贝叶斯分类器在电力系统负荷预测中的应用

摘要

Security and stability load forecasting of power system play a very important role. Factors of affecting power sys-tem short-term load forecasting are mainly environment, such as temperature, sunshine and humidity. These factors are com-plex. Considering all the factors will cause information redundancy and reducing prediction precision. Bayesian classifier im-proved author can be effectively processed all kinds of factors that affect load variation of power system, and exported its core factors. On this basis, accuracy of power system load forecasting was greatly improved. The authors forecasted load of a area by using the method. The results showed that the method had advantages of effectiveness and accuracy.%负荷预测对电力系统的安全稳定有十分重要的作用。影响电力系统负荷短期预测的主要因素是环境，如温度、日照、湿度等，这些因素关系复杂，综合考虑所有因素会导致信息冗余，降低预测精度。笔者改进的贝叶斯分类器可以有效地对影响电力系统负荷变化的各种因素进行处理，导出其中的核心因素，以此为基础大大提高电力系统负荷预测的精度。应用此方法对某地区负荷进行了预测，结果表明该方法的有效性和准确性。