Allaoua BrahmiaAdlen KerbouaRidha KelaiaiaAmeur Latreche2024-05-262024-05-262023-07-26https://www.mdpi.com/2076-3417/13/19/10703#http://dspace.univ-skikda.dz:4000/handle/123456789/1858Robotic systems require high accuracy in manipulating objects. Positioning errors are influenced by geometric tolerances and various sources. This paper introduces a new technique based on the interior-point algorithm optimisation method to allocate tolerances to the geometric parameters of a robot. This method consists of three steps. First, a method for modelling the kinematic problem as well as the geometric errors must be used. The Denavit–Hartenberg rule is the most suitable method for this modelling case. Then, a mathematical model for tolerance allocation is developed and used as a nonlinear multivariable optimisation problem. Finally, the “interior-point” algorithm is used to solve this optimisation problem. The accuracy and efficiency of the proposed method, in determining the tolerance allocations for a Delta parallel robot, is illustrated via calculation and simulation results. The values of the dimensional tolerances found are optimal. As a result, these values always keep the accuracy less than or equal to the imposed value.enArticle