The scheduling function in flexible manufacturing cells.

Blaga, Florin Sandu

Abstract: The Computer Aided Programming and Scheduling (CAPS) is an important attribute of modern flexible manufacturing. This facility has a lot of functions regarding production planning level, and these functions must be integrated with the others facilities by computer. The paper presents the Production planning function for Flexible Manufacturing Cell case- FMC-2R-2002 from University of Oradea. The simple priority rules will be used in decision procedures based on fuzzy sets. For each schedule manufacturing possibility the Gantt graph and efficiency indicators will be determinate. The decisional unit chooses the most adequate alternative for the considered objectives. After manufacturing program selection is done, the FMC-2R-2002 command program is automatically generated

Key words: flexible manufacturing, fuzzy sets, scheduling.

1. INTRODUCTION

The planning, programming and control affects FMS performances. The manufacturing scheduling is components part of these activities.

In most cases the manufacturing schedule is a decision resulting, based on certain pre-establish rules and is focused on some objectives. The schedule is materialized on time system resources allocation.

The demands of modern manufacturing have as results development a many scheduling methods. A classification of scheduling methods is presented in (Starbek et al., 2001). Decisional techniques based on fuzzy sets are used to define schedule-manufacturing procedures in FMS (Pandian, 2006), (Politano et al., 2001).

Using of mathematical modelling (Brucker et al., 2006), at manufacturing scheduling allows obtainment of optimal values for the considered criteria. This advantage is decreased by difficulties in model development, difficulties that primarily are generated by the necessity to include some system specific restrictions in the model.

Scheduling procedures for manufacturing presented in different references are described in most cases from the point of view of priority establishment (simple priority rules, fuzzy techniques, mathematical modelling). For a production task the part manufacturing order is established, this is then transformed in a manufacturing schedule which is implemented off-line in the manufacturing system. A dynamic programming (scheduling) procedure of the manufacturing process used in a flexible manufacturing system is presented in (Shnits et al., 2004).

Departing from this observation the paper presents an integrated programming system which has the following functions:

A. Takeover and processing of information regarding the parts which forms the manufacturing task. The result of information processing yields the organization of this information in a form that is proper to manipulate, as inputs in the fuzzy sets based decisional process.

B. Decision process performing for different input information in order to establish part manufacturing priorities in the manufacturing structure (flexible manufacturing cell, flexible manufacturing system).

C. Determination of efficiency indicator values for every scheduling variant and delivery of values to the decision factor (user) in order to be analyzed.

D. Comparison of different scheduling variants concerning a certain efficiency indicator.

E. Performing objectives C and D the decision factor (the user) is offered the necessary information in order to select one of the scheduling variant.

F. Automated generation of control and monitoring program of the flexible manufacturing system (cell). This program commands the operation sequences which are to be executed by the system's components so that the manufacturing task can be fulfilled.

G. Evaluation of flexible manufacturing cell operation by modelling and simulation.

H. Implementation of manufacturing program in order to execute the manufacturing task.

2. THE SCHEDULING FUNCTION

The steps of manufacturing scheduling (the principle of scheduling function) in FMC-2R-2002 cell are presented in figure 1.

1. Efficiency criteria definition on which the different scheduling variants will be compared;

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

2. Input of part specific parameter values so that can be organized in different input information sets for the decisional procedure. The parameters are following: arrival time, [R1.sub.i]; due (delivery) date, [d.sub.i]; the corresponding amount of profit for the ith product, [prof.sub.i], due (delivery) date penalties corresponding to the itch product, [p.sub.i]; total processing time, [TT.sub.i]; a surface processing time, TT[1.sub.i]; B surface processing time, TT[2.sub.i]; Number of tools used in manufacturing process NR_[SCULE.sub.i], where i = 1, ..., n.

This information is used as input in an 8x8 matrix, called Input Data Matrix (IDM).

3. Simple priority rules establishment which will be used in the decisional procedure, having a correlation between these rules and the part specific parameters. The simple priority rules identification that will be used in decision procedures:

R1: the product with the SDD (Shortest Due Date) will be selected;

R2: the product with the minimum static slack time will be selected; R3: the product having the minimum ratio between the maximum time interval in which the; product has to remain in system and the entire process interval will be selected

R4: the product having the maximum total processing time will be selected;

R5: the product with the maximum number of tools used in manufacturing process will be selected.

4. Generation of crisp input values matrixes. These matrices will contain characteristic values for every part from the manufacturing task, grouped so that they could be used to determine manufacturing priorities with the help of a lot of decisional system variants.

5. Performing of decisional procedures based on fuzzy sets in different defined variants. The decisional process, based on fuzzy sets supposes the development of an inference engine (rule base). The rules which are composing the inference engine will be defined by aggregation of a lot of simple priority rules using fuzzy set specific operators.

6. Manufacturing priorities obtainment of parts for different decision procedure variant run.

7. Establishment of manufacturing scheduling variants (part manufacturing steps) in the manufacturing system.

8. Automated generation of GANTT graph and calculus of performance indicator values for every variant of manufacturing scheduling.

9. Considering every defined efficiency criteria, the decision factor can compare and evaluate every scheduling variant, having the possibility to select the optimal one according to it's demand.

10. For the selected scheduling variant it is automatically generated the flexible manufacturing cell control program. Realization of this program has at its base the part manufacturing order established by the decisional procedure. Practically, the control and monitoring sequences are inserted following a logic imposed by the selection made by the decision factor (the user).

11. For the automated control and monitoring program generation a data base will be developed which will contain all the sequences of the program at a given time.

12. Flexible manufacturing system operation for the adopted control program can be verified by simulation using a Petri nets model.

13. An evaluation of modelling and simulation results will be made. Depending on this result, the decisional factor establishes a loop to one of the previous steps, iteratively, until the evaluation will give the best results.

14. Implementation of the generated control and monitoring program in the flexible manufacturing cell.

3. CONCLUSIONS

The paper presents the development method for real manufacturing scheduling function in MFC-2R-2002. The characteristics of this function are the following:

--fuzzy techniques adaptation to specifically decisional problems of manufacturing priorities determination;

--many manufacturing program variants are proposed for a manufacturing task. Different efficiency indicators characterize each variant. Analyzing these indicators the decisional unit can choose the most adequate alternative for the considered objectives;

--the chosen manufacturing program has as correspondent an automatically generated cell command program. - the adopted program is validated by Petri nets simulation. All these principal attributes are integrated in software package that can be easily accessed by decisional factor.

The future researches will focus on complex structures FMS scheduling function development. Also we will develop the fuzzy decision system for better accuracy of outputs.

4. REFERENCES

Brucker, P., Knust, S. & Oguz C. (2006), Scheduling chains with identical jobs and constant delays on a single machine, Mathematical Methods of Operations Research, Vol. 63, No.1, pp. 63-75, ISSN 1432-2994

Pandian, M. V. (2006), Fuzzy production planning and its application to decision making, Journal of Intelligent Manufacturing, Vol.1, No. 17 (February, 2006), pp. 5-12, ISSN 0956-5515

Politano, P. R., Kato, E. R. R., Morandin Jr, O. & Camargo, H. A. (2001), Automatic Manufacturing Systems Scheduling Based on Fuzzy Logic, The 5th World Multiconference on Systemics, Cybernetics and Informatics SCI 2001, Orlando, USA

Shnits, B., Rubinovitz, J. & Sireich, D. (2004), Multicriteria dynamic scheduling methodology for controlling flexible manufacturing systems, International Journal of Production Research, Vol. 42, No. 17 (September 2004), pp. 3457-3472, ISSN 0020-7543

Stabek, M., Kusar, J. & Brezovar, A. (2001), Optimal Scheduling of Job in FMS, The 34-th CIRP International Seminar on Manufacturing, Atena, Grecia