Abstract: In order to scientifically and accurately forecast decoding amount of railway, this paper puts forward a kind of grey particle swarm combination model for accurately forecasting decoding amount of railway based on mutation particle swarm optimization algorithm and grey prediction method, where this algorithm is able to optimize the parameters and this method is able to accurately predict the system with the uncertainty factors. This paper also uses examples to analyze the prediction precision and feasibility of the grey particle swarm combination forecast model and compare with the traditional grey forecast model, the results show that this model obviously is better than the traditional grey forecasting model. Using this model to forecast railway future amount of railway is able to reasonably prepare the railway marshalling station and inspect the plan of operation, thus it provides a theory basis for marshalling yards planning and designing.

Key words: grey forecasting, mutation particle swarm optimization algorithm, marshalling station, decoding amount

摘要： 为了科学准确地对铁路解编作业量进行预测，基于变异粒子群算法优化参数的良好性能和灰色预测法对不确定因素影响的系统准确预测的优点，提出了一种灰色变异粒子群组合预测模型，对铁路解编作业量进行准确地预测。并通过实例分析了模型的预测精度和可行性，且与传统的灰色预测模型进行比较。结果表明，灰色变异粒子群组合预测模型对铁路解编作业量预测明显优于传统的灰色预测模型。运用该模型预测未来铁路的解编作业量，以对铁路编组站进行合理编制和检查运营计划，从而为编组站规划和设计提供理论依据。

关键词: 灰色预测, 变异粒子群算法, 编组站, 解编作业量