In the context of enterprise digital transformation, ERP/MES/PCS and other information systems have achieved phased results, and system integration has become an inevitable trend. Among them, MES is the hub of the planning layer and the workshop control layer, which has the production scheduling function with ERP, while ASP is the advanced production scheduling system, which has a more powerful and detailed algorithm. APS, whether as an independent system or as a component of MES or ERP, should straighten out its data flow, so this paper expounds the data integration and implementation methods of APS, MES and ERP.

I. Introduction of MES and APS

MES production execution system is a set of production information management system for workshop execution level of manufacturing enterprises, located between ERP and PCS process control system of enterprise resource planning layer. As an intermediate bridge between ERP and PCS, MES helps enterprises realize vertical integration and opens up the information flow from equipment, production to planning layer, and has been widely promoted and applied in enterprises.

The Manufacturing Execution Systems Association (MESA) has summarized 11 main functional modules covered by MES, such as detailed scheduling of processes, resource allocation and status management, production unit allocation, process management, human resource management, maintenance management, quality management, document control, product tracking and product inventory management, performance analysis and data collection. The actual MES products are usually customized according to the enterprise development, combined with the business model and demand characteristics, select several modules to form an MES system to support the operation of the enterprise. In the eleven functional modules of MES, detailed scheduling of processes, resource allocation and state management and production unit allocation are closely related to the APS system described below.

APS Advanced Planning and Scheduling System is a professional enterprise scheduling management software. It is a synchronous, real-time, constrained simulation capability for all resources, whether materials, machinery, equipment, personnel, supplies, customer demand, and other planning factors. APS system mainly solves the problem of accurate prediction of delivery capacity, detailed planning of process production and optimal material supply under the condition of limited capacity. APS system must be used to solve complex production planning and scheduling problems. APS system can not only make a reasonable and optimized detailed production plan, but also combine the performance with the plan. It is an effective tool for enterprises to realize lean production and agile manufacturing. For complex scheduling problems under multiple constraints, APS core algorithms play an important role, and the main algorithms applied to APS are heuristic algorithm and intelligent optimization algorithm.

Heuristic algorithms include: heuristic rule method, expert system method and Lagrange relaxation method; Intelligent optimization algorithms include neural network (NN), simulated annealing (SA), Genetic algorithm (GA) tabu search (TS) and particle swarm optimization (PSO). These algorithms have their own characteristics for some specific needs, some "calculate fast", but the result is not the optimal solution, and some convergence is extremely slow and impractical. For APS mathematical solution problem, the American applied mathematician EYUANSHI proposed the segmentation nesting (NP) algorithm, proved to generate Markov chain, achieve global convergence, and can give the confidence interval from the optimal solution, which is recognized as an effective algorithm to solve the optimization of complex systems.

The relationship between APS and MES and ERP is shown in Figure 2 below. PCS system is located at the bottom of the system for the purpose of integrated management and control of equipment, MES is located between ERP and PCS, and APS overlaps with MES and ERP. If APS is an independent subsystem, its function can not be realized without the integration of ERP and MES data. In the development process of APS, it was once included in the ERP manufacturers, and now it is gradually closer to MES, some MES products provided by MES manufacturers are equipped with APS scheduling functions, in view of the powerful algorithm of APS, it is also sold as an independent information system product on the market.

From this process, we can also know the close connection between APS and MES, ERP system. Therefore, it is necessary to study what data interaction is involved between systems and how to achieve integration between systems. The benefits of APS and MES integration are also obvious, which can maximize the resource efficiency of the workshop, timely deal with changing orders and changing resources, quickly solve the problem of inserting orders, reduce downtime and waiting time, ensure the reasonable allocation of inventory materials, and realize flexible scheduling through intelligent scheduling.

2. Integrated data analysis

(1) Planned resource data

Planning resource data is mainly related to enterprise operation data, such as ordering contract, sales contract, delivery date, logistics period, etc. For ERP, APS, planning resource data is the source of the system's business, but also the data basis, is the first part of the demand into data, information data, available for the system to use. In fact, the MRP algorithm in ERP is to electronize the usual manual work mode, and it is also the basis for the formation and circulation of planning resource data. By integrating planning resource data, APS can form a master plan, which can then be refined into workshop plans, operation plans, etc.

(2)BOM data

BOM can be divided into three categories: design BOM, process BOM and manufacturing BOM. Generally, the design department in the enterprise designs products according to the order, forming the design BOM; According to the design BOM provided by the design department, the process department completes the process preparation, material quota, process approval, etc., and completes the flow parts splitting and process BOM output through the ERP interface program; According to the process BOM, the field personnel complete the tooling configuration, material distribution, process route, process regulation output, and output the manufacturing BOM through the ERP interface. In this ERP MRP can decompose the BOM calculation, APS system calculation performance is stronger, slowly replace the MRP calculation work, develop a more detailed BOM. MES also runs on manufacturing BOM data.

(3) Material data, etc

Material data, including inventory data and work-in-process flow data, are usually managed by MES through barcode system, RFID, manual entry, etc., and basic material information can be entered and unified management in advance in the MES database. MES integrates PCS layer data to provide reliable data information for ERP and APS.

3. Integration technology realization method

In technology, there are generally three methods of data integration: using the same database table; Use public documents; Automatic transfer of program or software data between databases.

(1) Use the same database form for the public part

Use the same database system, such as Oracle, Sqlserver, etc. The BOM table can be considered to be the same table. However, due to the closed nature of information system manufacturers, this approach is often limited. This integration approach can be taken if it is between the same manufacturer or a better open system.

(2) Using automatic data transfer methods

Automatic data transfer, is to transfer data through the database trigger, transmission is easy to achieve, low cost. Sql server is a commonly used database in APS, MES and ERP systems. Sql server database with trigger function, the use of trigger function to write code, code is automatically triggered. It is activated immediately after any changes are made to the data in the table, and then the program runs to make changes to other data tables. There are three basic types of triggers :Ainsert trigger: Triggered when data is inserted into a table; Bupdate trigger: Triggered when modifying data in a table. Cdelete trigger: Triggered when data is deleted from a table.

(3) Use middleware

In order to integrate heterogeneous systems and share information resources in digital manufacturing integration environment, the concept of middleware MW is proposed. Middleware is a piece of software that shields underlying heterogeneity by specifying a series of application programming interfaces that transform access and connectivity between the client and the server. Database communication refers to the storage of the machining quality data and process data of the current workpiece in the specified database form opened to MES by the field equipment directly through the control program.

(4) Use XML

Data exchange Standard Extensible Markup Language XML Unified Product Information Model. ERP/PDM/MES/PCS database system is a typical distributed database. The database is distributed in management level, workshop or production unit, which makes it difficult to integrate and share information in distributed heterogeneous database system.

XML is a simple and flexible standard format, which is suitable for the collaborative work of online publishing and data exchange to cross-platform cross-region heterogeneous applications, and meets the information integration and sharing requirements of distributed heterogeneous database systems. Using XML between different databases can use the defined common data structure to exchange information, but also can customize the format, the receiver and receiver through XML to achieve the purpose of information sharing and data exchange. Mainly used for MES and ERP system integration.

Iv. Summary

In view of the APS, MES, ERP system integration may involve different system suppliers, it is recommended to take the database intermediate table method, independent development is completed, in order to ensure higher integration development efficiency at the same time, define the database intermediate table form, clear data recognition.

APS can be sold as an independent system or integrated in MES to support the scheduling algorithm of MES. Studying the data flow integrated between MES and APS, including the data interacting with ERP, can help enterprises understand the source and flow process of their integrated data in the integration of information systems. According to the specific enterprise process, data integration should be comprehensively considered, APS should be attached importance to the application of APS in the business, and appropriate APS and MES system integration implementation methods should be adopted to provide information system construction ideas for the digital transformation of enterprises.