Industrial performance concerns numerous criteria, often in interaction and of complex nature, not related to one elementary measure. Performance Measurement Systems (PMSs) have been developed to support decision-making for reaching the objectives and launching adequate action plans. PMSs provide thus performance expressions which identify objective satisfaction degrees. Two kinds of performance expressions are useful in industrial problems, according to the scale (unipolar, bipolar) that is used for their definition. Moreover, these expressions generally have to be synthesized for global control purposes, determining an overall performance raises the issue of performance aggregation. To address such an aggregation issue, adequate multi-criteria methods need to be implemented. Most of the approaches proposed in the literature either do not provide explicit mechanisms, or rely on too simple methods. This paper deals with the definition of a performance combination based on mathematical tools, especially the generalized Choquet integral to take into account on the one hand criteria interactions and on the other hand both unipolar and bipolar scales. An application to a PMS for the service rate of a SME producing kitchen elements is used to illustrate the approach.