DE19649643A1 - Maintenance strategy method for machine or plant - Google Patents

Abstract

The maintenance strategy method uses a decision logic with successive hierarchy planes, for indicating the need for fault maintenance, timed maintenance and operative maintenance. The decision logic has a first hierarchy plane (1) in which yes/no decisions are provided in response to simple criteria. A second hierarchy plane (2) is provided, in which decision criteria are multiplied by point weightings and used for evaluating the prior decisions. A third plane uses a fuzzy logic for deciding the criteria for the first hierarchy plane.

Description

The invention relates to a method according to the preamble of claim 1 Device according to the preamble of claim 2 and a connection to Databases of operational management systems according to the preamble of claim 3.

Applications exist in the entire area of machine and maintenance Investments. The procedure allows, based on the existing level of State knowledge of the components, the assignment of the optimal conservation type (e.g. Maintenance), the decision making depending on the existing case Data and knowledge base, on the first level relatively quickly through yes / no Decisions, if necessary, but also by evaluating differentiated criteria subordinate levels can take place. Orient the decision branches the possibilities, the technical and business condition of Include assemblies and components in the strategy selection. The structure of the first level is based on assignments of possibilities of state determination defined types of repair / 1 /.

There are extensive efforts for the selection of maintenance strategies for Develop machines and systems processes / 2 /, / 3 /, / 4 /.

The task of your own solution is through a new hierarchical Structure of the process in several levels, the selection of maintenance strategies knowledge-based, based on the existing state knowledge about the assembly make that the effort of strategy selection greatly reduced and compared previously known methods are individually adapted to the individual case.

This object is achieved in a method according to the preamble of claim 1 in that

• - that the fault repair, the time-related repair or surgical repair may be selected directly in accordance with Figure 1 on the first three levels of the hierarchy by simple yes / no decisions.
• - That according to the example of FIG. 2, if necessary, a characteristic (P) is formed on a second level by rating differentiated criteria, multiplying the ratings (B _i ) by weightings (G _i ), standardization and subsequent summation over all criteria can be used for decision making in the first level and
• - that on a third level with the help of artificial intelligence (e.g. fuzzy logic) a decision can be made

it is solved in the device according to the preamble of claim 2,

• - that decision criteria and strategies of the first hierarchy level as fields, Decision paths through connecting lines between the fields and Decision options through buttons on the connecting lines as graphical user interface are displayed, so that by pressing the Buttons on the decision criteria of the assembly a strategy can be assigned
• - That with a button "decision aid" in the fields window of the subordinate second level can be opened, from there Sub-criteria of explanatory sub-criteria a characteristic value (P) calculated and compared with a limit value, so that the characteristic value P on the first level, if it is greater than the limit, at the "yes" or if it is less than the limit value at the "no" button of the associated Decision criterion is displayed and thereby a proposal for the Decision of the level 1 criterion arises,
• - That by pressing another button window of the third Hierarchy levels are opened, in which with the help of artificial intelligence (e.g. Fuzzy logic) a proposal for the decision at the first level arises

and it is solved by connecting the device according to the preamble of claim 2 to databases of operational management systems,

• - That the module name together with the assigned Repair method and the chosen decision-making path for the purpose of Further processing can be stored in a database.

The basis for the decisions is the existing knowledge about the state of Assemblies and components. In the first level, the strategy selection is based on of general criteria through yes / no decisions. The second level represents differentiated sub-criteria ready for the decision on the superordinate criterion, the due to different importance for the decision with weighting factors documented and assessed by awarding evaluation points with regard to their fulfillment be, so that after the subsequent normalization by division by the maximum possible evaluation number and summation over all criteria a characteristic value for the Decision of the first level criterion can be formed. By definition a limit value for the parameter formed and by comparing the calculated value The characteristic value with the limit value provides decision support for the  Criterion of the higher level 1. In the third level, with the help of artificial Intelligence (e.g. fuzzy logic) found a proposal for the overriding criterion.

application areas

The procedure has been developed for the selection of maintenance strategies for Machinery and equipment. But it is also for similarly based knowledge-based ones Decision-making processes suitable. The software according to the preamble of claim 2 due to the simple structure also allows the inexperienced user of the system a problem-free application of the method according to the preamble of claim 1.

References

/ 1 / Sturm, A .:
Condition-based maintenance - experiences and problems VGB Kraftwerkstechnik 73 (1993), 12; Pp. 1028-1030
/ 2 / Warnecke, H.-J .:
Maintenance - basics
Publishing house TÜV Rheinland, 1981
/ 3 / Lauenstein, G .:
Maintenance strategies for machines and plants
Linde publishing house, Berlin 1993
/ 4 / Beckmann, G .; Marx, D .:
Plant maintenance
Publishing house for basic materials industry, Leipzig 1994

Claims (3)

1. Procedure for assigning assemblies and components to repair types, for extending the service life and utilizing the stock of wear, in which the assignment can be based on knowledge on three hierarchical levels in different depths, characterized in that the strategy selection on the first level based on simple criteria yes / no decisions can be made that on the second level by multiplying the point ratings of differentiated criteria with weighting, standardization and subsequent summation over all criteria, a parameter is formed as the basis for the decision of the overriding criterion and that in the third level with the help of artificial intelligence (e.g. fuzzy logic) a decision for the criterion of the first level can be found. 2. Device for performing the method according to claim 1 for the assignment of Modules for repair types, the method according to claim 1 graphically is depicted, characterized in that in the first level Decision criteria and strategies through fields, decision paths through Connection lines between the fields and decision options Buttons on the connecting lines are represented as a button "Decision support" in the criteria fields window of the subordinate hierarchy levels opens that in the second level from entered point ratings a differentiated sub-criteria a parameter (P) as the basis for a decision Support is calculated on the first level and displayed in the third An assignment can be made at the level through artificial intelligence (e.g. fuzzy logic). 3. Connection of the device according to claim 2 with databases of operational management systems from which assemblies are called and in which after the strategy assignment Assemblies with the selected strategy and the chosen decision path can be saved for the purpose of further processing.