Optimization of Controller Parameters of Rudder Roll Stabilization Using Simplex Optimization Algorithm

Considering the problem of optimizing controller parameters for ship rudder roll stabilization when the closed-loop gain shaping algorithm is used to design the controller, a simplex optimization algorithm is proposed. Specially, integral of time multiplied by the absolute value of error (ITAE) criterion is proposed as the performance index for simplex optimization algorithm, and the controller parameters are optimized on-line according to the output of actual system. The optimization of controller parameters are realized by the simplex optimization algorithm, and the results show that the efficiency of setting the control parameters is improved and the roll amplitude is significantly reduced. In addition, the negative influence of rudder roll stabilization to yaw can be also reduced. This method of simplex optimization algorithm provides a new theory for rudder roll stabilization and establishes a theoretical foundation for the application in practical engineering.

Composite Scarf Repair Patch Optimization Under Uniaxial Loading Conditions

Simplex optimization algorithm was applied to predict the fiber orientations of the scarf repair patch for a given quasi-isotropic panel to maximize strength retention of the repair. The optimal stacking sequence of the repair patch avoids 0 degree plies in the direction of the load. Such a stacking sequence prolongs the life of the adhesive and results in a predicted 13% strength increase as compared to the traditional ply-by-ply replacement. The second optimization problem solved was one of finding the least favorable stacking sequence of the repair patch. Such stacking sequence inserts stiff plies into the patch and leads to the failure of the repair patch as early as 20% below the reference failure load of the repair patch with traditional ply-by-ply replacement. The strength prediction model consisted of nonlinear constitutive modeling of adhesive behavior and fiber failure prediction loads in the adherents based on critical failure volume (CFV) (see [8]) strength prediction method. Benchmark analysis was performed on the virgin, scarfed, and repaired (ply-by-ply replacement) panels and was in good agreement with experimental data.