Research Of Electromechanical Leveling System Base On Multi Machine Communication

Automatic leveling control system has been widely used in military, civilian areas, study is made from these two aspects aiming improve the speed and accuracy of leveling system in this paper.Firstly, vehicle platform was taken as example to analyze the system requirements, select reasonable parameters, provide solution for the problem of virtual legs during leveling. Proves the convergence of the control algorithm through analyzing platform position. The models of supporting system are also established according to the previous parameters to provide the basic measurement comparing the efficiency of algorithms.Secondly, the similarities and the differences of control strategies are compared, and the best strategy which suits the self-leveling system will be selected. The BP neural network is introduced to make up the disadvantage of classic PID control, which is difficult to adapt to dynamic nonlinear uncertain systems.Finally, simulation model of the leveling system is established according to the previous system model and the control algorithm. Through the comparison of the experiment, it can be seen that the PID control based on the BP neural network can provide better robustness and adaptability than classic PID control, which achieve better control effect. The core parts of the control system, including the DC motor drive circuit and multi-machine communication protocol, are designed. Choses the WINCE6.0 operation system for the host computer and transplant the algorithm into program.