As can be seen in Table 16, the inter-cell material handling cost is zero. This means that all part types are processed in
only one cell and the value of decision variable Yikh for each part type and each period is 1 only once.
The processing of parts on machines along with the assigned workers for three periods are shown in Fig. 3. For instance,
part types 5 and 6 are produced in the same position and workers in any periods. Part families, machine groups, worker
assignment are also depicted in the cell configurations presented in Table 17. Based on this Table, relocation is occurred for
machine type 1 where it is added to cell 1 in period 2. One worker of type 1 is hired to cell 2 in period 1 and fired from the
system in period 3, and worker type 1 is hired to cell 2 in period 2. Moreover, in cell 2, one worker of type 1 is hired and one
worker of type 3 is hired in the second period.
The linearized proposed model consists of 524 variables and 1186 constraints in the first example and its CPU time
is 37 min. The second example consists of 1062 variables and 2475 constraints and its computation time is 980 min.
The proposed model is computationally complex as it integrates the dynamic cell formation problem along with other
manufacturing features including the cell reconfiguration and the part routing problem with alternate workers. The cell
formation problem has been reported as a NP-hard problem [25,26]. Logendran et al. [27] have shown that the problem of
the determination of the process routing from alternate routings is also NP-hard. Chen [10] described that solving the model
considering system reconfiguration in terms of machine relocation is NP-hard as well. Therefore, the proposed model in this
paper is NP-hard since it integrates all these NP-hard problems.
As can be seen in Table 16, the inter-cell material handling cost is zero. This means that all part types are processed in
only one cell and the value of decision variable Yikh for each part type and each period is 1 only once.
The processing of parts on machines along with the assigned workers for three periods are shown in Fig. 3. For instance,
part types 5 and 6 are produced in the same position and workers in any periods. Part families, machine groups, worker
assignment are also depicted in the cell configurations presented in Table 17. Based on this Table, relocation is occurred for
machine type 1 where it is added to cell 1 in period 2. One worker of type 1 is hired to cell 2 in period 1 and fired from the
system in period 3, and worker type 1 is hired to cell 2 in period 2. Moreover, in cell 2, one worker of type 1 is hired and one
worker of type 3 is hired in the second period.
The linearized proposed model consists of 524 variables and 1186 constraints in the first example and its CPU time
is 37 min. The second example consists of 1062 variables and 2475 constraints and its computation time is 980 min.
The proposed model is computationally complex as it integrates the dynamic cell formation problem along with other
manufacturing features including the cell reconfiguration and the part routing problem with alternate workers. The cell
formation problem has been reported as a NP-hard problem [25,26]. Logendran et al. [27] have shown that the problem of
the determination of the process routing from alternate routings is also NP-hard. Chen [10] described that solving the model
considering system reconfiguration in terms of machine relocation is NP-hard as well. Therefore, the proposed model in this
paper is NP-hard since it integrates all these NP-hard problems.
การแปล กรุณารอสักครู่..