One of the main problems of networked control systems is that signal transmission delay is inevitable due to long distance transmission. This will affect the performance of the system, such as stability range, adjustment time, and rise time, and in serious cases, the system cannot maintain a stable state. In this regard, a definite method is adopted to realize the compensation of network control system. To improve the control ability of mobile sensor network time delay system, the control model of mobile sensor network time delay system based on reinforcement learning is proposed, and the control objective function of mobile sensor network time delay system is constructed by using high-order approximate differential equation, combined with maximum likelihood estimation method for parameter estimation of mobile sensor network time delay, the convergence of reinforcement learning methods for mobile sensor network control and adaptive scheduling, and sensor network time delay system control model of multidimensional measure information registration in strengthening tracking learning optimization mode to realize the adaptive control of mobile sensor network time delay system. The simulation results show that the proposed method has good adaptability, high accuracy of estimation of delay parameters, and strong robustness of the control process.