Model Predictive Control (MPC) has become an important control method for autonomous vehicles and complex robotic systems. However, MPC requires solving an optimization problem to ensure optimal control inputs, which can be computationally expensive for nonlinear and high-dimensional systems. This paper proposes using Particle Filters (PF) to execute the solving process of MPC to enhance efficiency and accuracy. Our approach applies PF to solve quadratic programming problems and integrates it into the MPC framework. We investigate two specific applications: lane-keeping control for autonomous vehicles and control of a robotic arm mounted on a differential drive mobile platform. Experimental results show that using PF can effectively optimize MPC problems, significantly reduce computation time, and improve control accuracy, particularly in handling complex nonlinear systems in the mentioned applications. This paper demonstrates the potential of PF as an optimizer in MPC and suggests further testing of this approach in more complex control problems to verify its broad applicability and reliability.