基于普通累积法的原油管线油电损耗预测

影响输油成本变化的主要因素是油电损耗费用，文献利用灰色系统理论，以榆油管线每年的输量、相应的电耗和油耗为时间序列，建立了三元素非等间距GM（1，1）模型来预测管线油电损耗，但是其结果精度并不高，而且计算复杂，油损耗预测的相对误差范围是0．013％-16．86％，平均相对误差是6．69％；电损耗相对误差范围是1．76％-20．74％，平均相对误差6．64％。文献采用回归法进行预测，该法精度较文献高，但是该法的显著特点是所需要数据太多，计算复杂。本文采用累积法进行预测，利用文献的数据，通过计算结果可以看出油损耗预测的相对误差范围是1．55％-5．59％，平均相对误差是3．48％；电损耗相对误差范围是0．55％-1．91％，平均相对误差0．97％。精度明显大副提高，且该方法简单易懂。所需数据少。是一种较好的预测方法。

FORECAST OF OIL AND POWER LOSS OF CRUDE OIL PIPELINES BASED ON GENERAL ACCUMULATION METHOD

The main factor causing changes of oil transportation costs is the expenditures of oil and power loss.The Literature~(3]) makes use of grey system theory to established three-elemental non-equidistance model to forecast the oil and power loss of pipelines with the time series of the annual of oil pipelines,corresponding power and oil loss,but the precision of results is unsatisfactory and the calculation is complicated: the relative error range of forecast of oil loss is 0.013%-16.86% and its average relative error is 6.69%;while the relative error range of forecast of power loss is 1.76%-20.74% and its average relative error is 6.64%.The Literature~(2]) conducts the forecast in regression method,whose precision is higher than that of Literature~(3]).However,it requires too many data with complicated calculation.This paper conducts the forecast in accumulation method with data of Literature~(3]).The result is shown as follows: the relative error range of forecast of oil loss is 1.55%-5.59% and its average relative error is 3.48%;while the relative error range of forecast of power loss is 0.55%-1.91% and its average relative error is 0.97%,which presents that the precision is obviously and greatly improved.Furthermore,this method is easy to understand and master,which requires fewer data,so it is a good forecast method.