Search results

Differential sensor for linear displacement has been optimized in terms of linearity and sensitivity. The optimization of the sensor is carried out with respect to its characteristic displacement‐output voltage.

Response surface methodology and design of experiments have been successfully applied for sensor optimization. First, a full factorial experiment – computation of the quasi‐static electromagnetic field of the sensor using the finite element method – has been performed. Secondary model has been created on the basis of finite element results. Then this model has been optimized with respect to two criteria – linearity and sensitivity.

The mutual influence of both criteria has been studied. In this way an optimal trade‐off between linearity and sensitivity of the sensor was achieved.

Relatively small numbers of optimization criteria have been varied. The results can be further improved by adding additional factors to the study.

The results obtained improve characteristics of this type of sensor and make it low cost alternative to the high‐end linear displacement sensors in low to moderate accuracy applications.

The purpose of this paper is to solve the optimization problem for a permanent magnet linear actuator with moving magnet for driving a needle in a knitting machine. The optimization is carried out with respect to the maximal average force along the stroke.

The optimization factors are the dimensions of the permanent magnet and the pole length. For obtaining the objective function for each parameter combination, the static force‐stroke characteristic is computed. Three‐dimensional finite element analysis is employed for obtaining the electromagnetic force. The objective function is then approximated by secondary models on the basis of design of experiment. The optimization is carried out using sequential linear programming.

Optimal solution to the problem has been obtained and the force‐stroke characteristic of the optimal actuator is given.

The proposed approach employs sequential linear programming with move limits using trust region concept, 3D finite element analysis, design of experiments and secondary models for optimization of a permanent magnet linear actuator for individual needle driving in a knitting machine.