Solver Strategies
Following are details about strategies on the Solver options to guide the search-algorithm and help it find and optimize a schedule more effectively.
The Priority Batches field defines a sub-selection of batches within the selection made by the user. The list boxes present a list of predefined and named searches. These can be defined by saving a user-defined selection in or .
The set of batches that are going to be sequenced is defined as the intersection of the current selection (selected batches on the planning board) and the priority batches. Sequencing means that the solver tries different sequences for these batches only. Other batches in the current selection (that are neither in the set of priority tank batches nor in the set of priority process batches) follow the order imposed by the placement of the priority batches. The order of non-priority batches is determined by the order of the priority batches; when a sequence for priority batches is chosen, the order of non-priority batches is left unchanged. In most cases, priority batches will include or comprise the batches that form the bottleneck in a factory.
- Priority Process Batches
This list box shows a sub-selection of the process batches in the current user-defined selection. By default, all selected process batches are marked as priority batches; other selections may be defined through
by saving a named search.If you do not want priority process batches, select empty selection.
- Priority Tank Batches
This list box shows a sub-selection of the tank batches in the current user-defined selection. By default, all selected tank batches are marked as priority batches; other selections may be defined through
.If you do not want priority tank batches, select empty selection.
- Sequence Priority Batches
If this check box is selected, the Solver schedules the selected priority batches in a specific predefined sequence, rather than trying all possible sequences searching for a feasible or optimal schedule. Using a single sequence is useful to limit the search space, when the optimal sequence is known in advance or when the user wants to enforce some specific sequence. The list box shows a number of predefined sequences, possibly extended by a user-defined sequence.
This strategy includes backtracking on time and enabling resource allocation decisions.
- Strict Sequence (Do Not Fill Gaps)
If strict sequence is selected, the selected sequence will not be interrupted, even if that sequence contains gaps that are large enough for other batches to fit in.
- First Allocate then Sequence
By default, the Solver schedules each batch (piece by piece), setting its time and assigning a resource simultaneously. If this check box is selected, the Solver will first allocate all batches to some valid resource and then - given this allocation - determine the best sequence and time values for each batch. Setting the allocations in advance drastically limits the search space, so this strategy enables the Solver to find a solution much more efficiently. However, since not all allocations are considered, it may fail to find a feasible schedule even though one would exist. This option is probably most useful in case of complex overlapping resource groups or choice groups containing heterogeneous resources with intricate changeovers. This option allows you to optimize changeover.