Module partitioning study of specialized engineering services based on modular theory ——case on coal production firms

Outsourcing of specialized engineering services has become one of the routes to the innovation and high-quality development of the commercial model of coal production enterprises in China. Currently, although there is practice, no scale-up development has been achieved.In this paper, based on modular theory, under the condition of unknown expert weights, triangular fuzzy number method is used to construct a service unit association relationship evaluation model, and the correlation matrix is calculated. Then, a genetic algorithm is used to cluster the association matrix and obtain the results of the division of the outsourced module of engineering services. Finally, the proposed method is validated using the coal mine development process of coal producers as an example.


Introduction
In 2016, the tenth five coal industry development program encouraged coal enterprises to establish specialized service companies [1]. This program provides specialized engineering services to coal production firms and provides policy support for conducting business model innovations.
Coal specialized engineering services are available at home and abroad [2]. Cum according to the service demand party needs, improve mine design, resource distribution, business management and other engineering service activities. However, in practice, specialized engineering services for coal production fail to achieve scale effects because service content is formed based on negotiation between the supply and demand parties, and as a result, there is a lack of standardized service products.
Hongling Xu (2007) argues that modularity in service product design is the basis of industrialization [3]. Modular theory was proposed by Baldwin and Clark at Harvard Business School in 1997 [4]. Xia Hui (2010) considers service modularity as the modularity of service processes, thus forming a series of activities to meet the demands of different markets [5]. Gomes (2008) and Zhang Qianjun (2015) argue that appropriate task modularity has a positive effect on firm performance [6] [7]. Based on the above studies, this paper argues that the service modularity theory can guide the coal production specialty to provide services for industrialization and scale development.
The basis of service module design is module partitioning. References [8][9][10][11][12][13][14] have studied the problem of task module partitioning from different perspectives and methods. However, there is a lack of multidimensional analysis of association relationships, in terms of quantitative research, neglecting the effect of expert opinion on evaluation results, and often using fuzzy clustering, the accuracy of its calculation results is affected by artificially set thresholds and is not objective enough. Therefore, this paper proposes a service modularity design method based on multidimensional fuzzy association relationship analysis and solving the optimal module partition scheme.

MODELING THE RELATIONSHIP BETWEEN SERVICE UNITS
The smallest unit studied in this paper is the service unit. It has independently complete functions, including the operation of accomplishing active tasks. Basis of module division.

Linguistic Scale of Association Mode and Relationship Strength
The relationship between i s and j s is general The relationship between i s and j s is

Ruantification of Service Unit Relationship
(1) Representation of service unit association matrix The relationship between service units is a binary relationship, which can be expressed as： l m u l m u l l m m u u a a a a a a a a a a a a a a   (2) Consistency of expert opinions and weight of experts Based on the evaluation information of each expert, the mean value of triangular fuzzy number of the strength of the relationship between service units is obtained The individual weight of expert P can be normalized by (7) (3) Service unit Association calculation The triangular fuzzy number of the association relationship is calculated according to the expert weight.
The average area method was used to make the image clear.

Construction of Objective Function
According to the literature [15] [16] in this paper, modularity degree is used to evaluate the module division results.

Identification Of Specialized Service Units Of Coal Production Enterprises
Based on the principle of technology behavior, combined with field investigation in coal production enterprises, this paper forms a list of service units as in Tab

Relationship Measurement of Professional Service Unit
Based on [2] and [17][18][19], this paper reviews the relationship between service units in three dimensions, safety, function and economy. Space is limited, only some of the results of this review are truncated, as summarized inTable4.

Table4. SEVICE UNITS SECURITY RELATIONSHIP EVALUATIO
The security association matrix of service unit is

Division Scheme Solution and Result Analysis of Coal Specialized Service Module
In this paper, the parameters of the GA were set as follows: n = 60, pc = 0.45, pm = 0.01. Calculations were performed using MATLAB 2016a version. The results showed that it could be divided into 14 modules, as shown in Table 5.

CONCLUSION
This paper presents the division method of engineering specialized service module and practices in combination with coal production enterprises. In cases where the expert weights were unknown, a triangular fuzzy number method was adopted to evaluate the association relationship and calculate the expert weights. And genetic algorithm is employed to calculate the optimal service module partitioning scheme. The research results of this paper provide the foundation for the development of the professional division of labour and professional service industrialization of coal enterprises.