基于BA-SVM及大数据的物资信息价预测

Material Information Price Forecasting based on the BA-SVM and Big Data Statistics

作为指导电力工程物资预算的信息价,为使其设计合理,需要结合工程物资价格的市场变动规律以及众多价格影响因素,而价格变动频繁,导致预测难度和数据分析工作量大。针对这一问题,使用蝙蝠算法-支持向量机预测模型,一方面能够通过蝙蝠算法提高支持向量机的预测精度,另一方面能够引入价格的影响因素对信息价进行预测,最终提高信息价的预测精度和解决大数据样本统计信息价的缺陷;然后以工程设备购置费用的结余率为造价控制指标,对信息价进行优化;最后通过案例分析验证了所提预测方法具有良好的适用性和实际使用价值。

As an information price of guiding power engineering material budget, to make its reasonable design, it requires a combination of engineering material price change rule of the market and many price influence factors, that lead to a big workload to data analysis and predict difficultly. In order to solve this problem, use the bat algorithm-support vector machine forecasting model, on the one hand, it can improve the prediction accuracy of support vector machine（SVM） with the bat algorithm, on the other hand can in-troduce price influence factors to forecast the information price, eventually improve prediction accuracy of information price and solve the defect of statistics price in large data sample; and then use the cost savings rate of engineering equipment purchase for cost control target, optimize the information price. Finally through the case analysis to verify the effectiveness of the proposed prediction method has a good applicability and practical significance.