Enhancing the Efficiency of Solving Global Dynamic Optimization Problems by Improving the Calculation of State Relaxations