The design of shaping pulse fields for controlling molecular orientation is of great importance in fields of stereochemical reactions, strong-field ionization, and quantum information processing. Traditional quantum optimal control algorithms typically solve the problem of molecular orientation in an infinite-dimensional rotational space, but they often overlook the constraints imposed by experimental limitations. In this work, a multi-objective and multi-constraint quantum optimal control algorithm is proposed to design a pulse field that conforms to the constraints of pulse area and energy. Specifically, the algorithm enforces a zero pulse area condition to eliminate the static field components and maintains constant pulse energy, ensuring compatibility with realistic experimental setups. Under these constraints, the algorithm optimizes the population and phase distribution of a selected number of low-lying rotational states in ultracold molecules to achieve maximum molecular orientation. The effectiveness of the proposed algorithm is demonstrated through numerical studies involving two- and three-state target subspaces, where the creation of a coherent superposition state with optimized population and phase distribution leads to the desired molecular orientation. Furthermore, its scalability is validated by applying it to a more complex 17-state subspace, where a maximum orientation value of 0.99055 is obtained, approaching the global optimal value of 1. Our findings demonstrate that by effectively managing these constraints, the influence of rotational states in the non-target state subspace can be substantially suppressed. The time-frequency analysis of the optimized pulses, combined with the Fourier transform spectrum of the time-dependent degree of orientation, indicates that the maximum molecular orientation is mainly achieved through ladder-climbing excitation of multi-color pulse fields, with the contributions from highly excited states being minimal. This work provides a valuable reference for designing experimentally feasible pulse fields using multi-constraint optimization algorithms, which helps to precisely control a limited number of rotational states to achieve maximum molecular orientation.