微型电脑应用

为了确保送电质量，提升供电可靠性，通过负荷预测为电力企业的规划策略提供数据依据，从电力负荷数据特性着手，详细分析了影响因素以及预测难点。采用PSO算法对LSTM长短期记忆神经网络进行参数寻优，并综合考虑温度、时段、电价等因素构建电力负荷数据预测模型；采用C#语言设计了包含数据仓库、模型构建与择优、负荷预测模块的C/S模式负荷数据预测平台。通过实际数据验证，预测精度较高，误差在允许范围内，为电力系统负荷数据预测提供了可靠的信息化手段。

In order to ensure the power quality, improve power supply reliability, and provide data through the load forecasting for the electric power enterprise planning strategy, this paper makes a feature analysis of power load data, detailed analyzes of influencing factors and predicts the difficulty, uses PSO algorithm to both short-term and long-term memory neural network tooptimize parameters by considering factors such as temperature, time, electricity prices, and to build power load data forecast model. The C# language is used to design the load data prediction platform of C/S mode, which includes data warehouse, model building and optimization, and load prediction modules. Through the example data verification, the prediction accuracy is high and the error is in the allowable range, which provides a reliable information method for the power system load data prediction.