This research introduces Intelligent Car Learning for Mobile Robot Path Planning (ICLMRPP). The primary objectives of these techniques are to expand the best minimum path distance, minimize energy consumption, ensure a smooth trajectory, minimize task completion time, reduce identification risk, enhance communication reliability, prioritize missions, maximize obstacle avoidance, and improve environmental sensing. The robot operates in both static and dynamic environments, utilizing a grid map. The holonomic and non-holonomic constraints related to car learning mobile robots are considered. The Multi-objective Deep Q Network (MODQN) algorithm is proposed to address the ICLMRPP problem. The obstacle avoidance, implemented using DQN in the proposed MODQN, incorporates nine improvements. The proposed algorithm is compared to the Multi-objective Double Deep Q Networks (MODDQN) technique. Experiments on a car learning mobile robot validate the efficacy of the proposed algorithm techniques. This work demonstrates that the proposed MODQN algorithm, when applied to ICLMRPP problems, generates safe paths with collision avoidance, optimized path distance efficiency, and practical implementability. The simulations of MODQN and MODDQN are deemed acceptable, with simulation and experimental deviations less than 3%, and a path distance error of 1.125%.