基于ABC的GA-NN模型在民机客服成本预测中的应用

发布时间：2018-05-13 05:00

本文选题：民机客户服务体系 + 民机客服成本预测　；参考：《南京航空航天大学》2012年硕士论文

【摘要】：近年来我国经济持续发展的同时，中国民航工业也得以迅猛发展，民机客服成本预测对我国民航工业的发展具有重要意义。如何对民机客服成本进行准确预测、提高客户服务质量和效率、降低客户服务成本是我国民航工业发展中亟待解决的问题。本文将作业成本法引入到民机客服成本核算中，提出基于ABC的GA-NN模型并将其应用到民机客服成本预测中，论文研究的主要工作如下：（1）提出了一种应用于民机客户服务体系的成本控制方案。该方案是一种多层次的，以作业成本法为核心、以成本分配观和过程分析观为导向，结合作业成本计算和作业管理的成本管理制度。（2）提出了一种应用于民机客服的ABC成本核算模型。该模型在原有成本核算功能的基础上，综合考虑了作业管理、业务流程优化、成本计算精度等多个方面问题，并将与服务相关的作业成本按照适当的动因分配给服务，使服务的成本信息更为准确，同时对服务的各项业务流程进行完整的描述。ABC成本核算模型不仅呈现民机客户服务体系业务流程中的成本走向和变化趋势，还能为GA-NN成本预测模型提供优秀的训练样本，它是保证GA-NN成本预测模型成功运行的关键技术。（3）提出了一种应用于民机客服的GA-NN成本预测模型。本文结合GA算法和BP算法的特点，重点阐述了GA算法全局优化NN网络的网络结构、初始权值和阈值的过程；在此基础上对NN网络进行训练，再运用双向确定法改变隐层节点数，不断对NN网络结构进行优化，，从而得到满足指定目标函数收敛条件的最优网络模型；最后利用得到的模型对民机客服成本进行预测。（4）通过将基于ABC的GA-NN模型和传统NN网络模型的预测结果进行比较分析，说明基于ABC的GA-NN模型预测的有效性。经实验分析后得到，基于ABC的GA-NN模型不仅能发挥GA算法强大的搜索能力，同时在预测精度和收敛速度方面也比传统NN网络模型更优秀。最后，本文将基于ABC的GA-NN模型的基本思想初步应用在民机客户服务作业成本管理系统中。应用结果表明该模型可以对民机客服成进行准确核算和预测。
[Abstract]:In recent years, while China's economy has been developing continuously, China's civil aviation industry has also developed rapidly. The cost prediction of civil aircraft service is of great significance to the development of civil aviation industry in China. How to accurately predict the cost of civil service, improve customer service quality and efficiency, and reduce customer service costs are urgent to be solved in the development of civil aviation industry in China. In this paper, the activity-based costing is introduced into the cost accounting of civil service service, and the GA-NN model based on ABC is put forward and applied to the cost prediction of civil service service. The main work of this paper is as follows:
(1) a cost control scheme applied to the civil aircraft customer service system is proposed. This scheme is a multi-level, based on activity based costing, cost allocation and process analysis, combined with the cost management system of activity based costing and operation management.
(2) a ABC cost accounting model applied to civil service customer service is proposed. On the basis of the original cost accounting function, the model comprehensively considers the problems of operation management, business process optimization, cost calculation precision and so on, and assigns service related operating cost to service according to appropriate motivation, so as to make the service cost information More accurate, the.ABC cost accounting model not only presents the trend of cost trend and change in the business process of the civil aircraft customer service system, but also provides excellent training samples for the GA-NN cost prediction model. It is the key technology to ensure the successful operation of the GA-NN cost prediction model.
(3) a GA-NN cost prediction model applied to civil service customer service is proposed. Combining the characteristics of GA algorithm and BP algorithm, this paper focuses on the network structure, initial weights and threshold processes of the global optimization NN network of the GA algorithm. On this basis, the NN network is trained, and then the number of hidden nodes is changed by the two-way determination method, and the NN net is constantly on the NN net. The structure of the collaterals is optimized to get the optimal network model which satisfies the convergent condition of the specified objective function. Finally, the model is used to predict the cost of civil service.
(4) by comparing the prediction results of the ABC based GA-NN model with the traditional NN network model, it shows the validity of the GA-NN model prediction based on the ABC. After the experimental analysis, the ABC based GA-NN model can not only give full play to the powerful search capability of the GA algorithm, but also be more accurate and convergent than the traditional NN network model. The type is better.
Finally, this paper applies the basic idea of the ABC based GA-NN model to the civil aircraft customer service operation cost management system. The application results show that the model can accurately calculate and predict the civil service customer service.

【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F562.6;F560.68

【参考文献】