To improve the comprehensive performance of vehicles equipped with stepped automatic transmission (SAT), the optimization of gearshift schedule should take into account various performance such as power performance, fuel economy, etc. In addition, the SATs would become more acceptable if the optimized gearshift schedule could also be individualized to reflect the driver’s expectation on vehicle performance to a reasonable extent. For the purpose of ensuring the comprehensive performance and improving the individual-ability (i.e., the ability to adapt to different driver’s performance expectation) of vehicles equipped with SAT, a linear weighted method has been proposed to construct a performance evaluation function, which applies different weights to represent driver’s expectation on performance by using these weights to multiply the normalized value of each sub-performance index. Taking this evaluation function as the optimization objective, several gearshift schedules of a vehicle equipped with an automated manual transmission (AMT) have been optimized with genetic algorithm (GA). Although the optimized power performance dominated and fuel economy dominated gearshift schedules can reflect the driver’s intention, the change of the control parameters is not very sensitive to the change of the weights in some areas. To this end, an improved evaluation approach to make the gearshift schedule be more sensitive to the change of the weights has been proposed. In this approach, the optimal solutions for each sub-performance are calculated at first, and then the square of the difference between the normalized optimal value and the normalized actual value of each sub-performance index is used to replace the normalized value of each sub-performance index in the linear weighted method, to construct the improved evaluation function. Simulation results show that the modified evaluation function can improve the sensitivity of the change of gearshift control parameters to the change of weights, and the gearshift schedules can be individualized by assigning different value to the weights to satisfy different driving preference.