Abstarct: Fixtures, as critical structural components in industries, are subjected to various loads, leading to elastic deformation and stress within their bodies. These deformations can negatively impact the performance of the fixtures. The positioning of locaters on the fixture body significantly influences their effectiveness. The main goal of this thesis is to develop a method for optimizing the placement of locaters on the fixture body to minimize the maximum elastic deformation of the fixture body. In this study, a Multi-laxyer Perceptron (MLP) Artificial Neural Network (ANN) was used as a predictive model for elastic deformation, and a Particle Swarm Optimization (PSO) algorithm was employed to determine the optimal positions of the locaters. Initially, 400 simulations were performed using Finite Element Analysis (FEA) software with various boundary conditions and loads on the locaters. The minimum deformation obtained in simulations was 0.06 for the fully constrained boundary condition and 0.12 for the pin boundary condition. Next, the ANN model was tested using the input and output data from the simulations. Subsequently, MATLAB code for the ANN was written, and the optimization algorithm was applied to find the optimal positions of the locaters, performed with 6 and 12 input parameters. FEA simulations were conducted for both cases to validate the optimal data. The simulation results showed that with 12 input parameters, the minimum deformation was 0.05 for the fully constrained boundary condition and 0.11 for the pin boundary condition. Experimental tests were conducted to validate the simulation results. For the experimental tests, a laboratory setup was first designed, followed by the manufacturing of components using various equipment and mechanisms. A total of 18 experiments were conducted, each with 3 repetitions. Forces were measured using load cells and applied to the fixture body plates. The deformation of the plates was measured using two dial indicators. Simulation results were compared with experimental results. For a plate subjected to two forces, the simulation result was 0.08 and the experimental result was 0.14, showing a difference of 0.06. For a plate subjected to one force, the simulation result was 0.08 and the experimental result was 0.10, with a difference of 0.07. Overall, the results demonstrated that the combination of the Artificial Neural Network-baxsed method and Particle Swarm Optimization algorithm is an efficient tool for optimizing the placement of locaters on the fixture body and minimizing its elastic deformation.