Designing MIMO controller by neuro-traveling particle swarm optimizer approach

A nonlinear, multiple input–multiple output controller called the quality controller of neuro-traveling particle swarm optimizer (QC/NTPSO) approach has been proposed in this paper. A reliable controller must stabilize the quality during the manufacturing process and bring the quality characteristics of the manufacturing process close to the target. This controller must also have an adequate feedback system with estimation technology and optimization algorithm. In addition, the artificial intelligence has reasonably been matured and is often used in dealing with construction problems. Therefore, this work constructed a controller with artificial intelligence technology by first using an artificial neural network as the predictor and then using the traveling particle swarm optimizer that is ideal for continuous optimization problems as the algorithm for optimization. The proposed approach has been tested through chemical mechanical polishing (CMP), an important process in semiconductor manufacturing. The result of the test shows that the proposed approach can bring quality characteristics closer to the target than any other approaches.