用自适应模糊推理系统预测电力短期负荷

为寻求有效的电力系统负荷预测方法以提高预测结果的准确度,提出了基于Takagi-Sugeno(T-S)模型的自适应神经模糊推理系统(ANFIS)。该系统采用减法聚类初始化模糊推理,把神经网络学习机制引入到逻辑推理中,并用混合学习算法调整前件参数和结论参数,自动产生模糊规则。考虑气象、日期类型等因素后将学习样本分为3组进行训练和检测。该方法对于受天气影响较明显的电网,能有效防止不合理预测结果的出现。对于武汉地区实际负荷的预测结果的分析表明该方法有较高的预测准确度,取得了令人满意的结果。

Short-term Load Forecasting in Power System Based on Adaptive Network-based Fuzzy Inference System

Adpative Network-based Fuzzy Inference System (ANFIS) is an effective method in the modeling of multi-factor load forecasting, however, the key step that influences the accuracy of forecasted results is the reasonable selection of the variables. ANFIS based on Takagi-Sugeno model is presented in this paper. The system of adaptive network-based fuzzy inference is initialized by subtractive clustering with nerve network is applied to fuzzy inference and adjusts the parameters of the fuzzy inference system with hybrid algorithm and can produce fuzzy rules automatically. An approach to automatically design the optimal fuzzy rule bases using ANFIS is proposed to construct the fuzzy models for short-term load forecasting. According to this approach, identification of the premise part and the consequence part is simultaneously accomplished, and the models complexity is also reduced compared to other fuzzy models. Considering the factors such as weather condition and type of day, train the data which are divided into three groups: training data set, checking data set and testing data set. This method can prevent unreasonable predict results, especially when the load of power system is remarkably affected by weather. Actually, we can reduce the error of load forecasting. The actual load forecasting results for Wuhan district show that the proposed method possesses better forecasting accuracy and the forecasting is satisfactory.