Study On Transfer Matrix Method For Dynamics Of Controlled Multibody Systems

In the field of weapon, aviation, astronavigation, ship, transportation, and general mechanical industry, many complex mechanical systems, such as multiple launch rocket system, self-propelled artillery, airplane, spacecraft, naval vessel, vehicle, robot, digit control machine tool, and so on, can often be regarded as controlled multibody systems including control system, rigid bodies and flexible bodies connected by different kinds of hinges. When using the ordinary methods to study the controlled multibody dynamics, it is necessary to develop the global dynamic equations of system, the orders of the involved system matrices are rather high for complex controlled multibody systems, and the computational time is very huge. Sometimes, it is difficult to satisfy the demand of rapid dynamics calculation and control system optimization. The transfer matrix method of controlled multibody system developed in recent years has the advantages as follows: without the global dynamics equations of system, high programming, low order of system matrix and high computational efficiency, etc. This method provides a new approach for dynamics design of controlled multibody systems. At present, the transfer matrix method of controlled multibody system is only used to study the dynamics of controlled multibody systems with certainty structural papramerts and known control laws.As one of the important research productions of national defence "973" project, in this paper, through the research on theory, calculation, and test, the transfer matrix method of controlled multibody system is extended as follows:(1) In aspect of the transfer matrix method of linear controlled multibody system, the algorithms of eigenvalue problems of linear multi-rigid-flexible-body systems is improved, and the repeated eigenvalues can be identified simply. The body dynamics equations and state space equations of linear controlled multibody system are presented, and the exact efficient dynamics analysis of controlled multi-rigid-flexible-body systems can be realized. (2) The finite element transfer matrix method of controlled laminate plates/shells is presented, and the speedy dynamics computation of controlled multidimensional structures is realized. (3) The H∞independent modal space control strategy is designed, and the active vibration control of linear multi-rigid-flexible-body systems and laminate plates/shells is realized. (4) In aspect of the discrete time transfer matrix method of controlled multibody system, the transfer matrices of some typical controlled elements are deduced, and the rapid dynamics calculation and highly efficient fuzzy neural network control of time-variable, nonlinear controlled multibody systems is realized. (5) The perturbation relationship between transfer matrices and random parameters of random linear multi-rigid-flexible-body systems and random laminate plates/shells is deduced. The eigenvalue problems of random systems are discussed, and the robust control design of random systems is realized. (6) The method presented in this paper is applied on two important engineering problems. Firstly, the control dynamics for the sealifts system of navy ships connected parallel is developed, and the PID adaptive controller of marine crane and fuzzy controller of wave compensation device are designed. The rapid dynamics calculation and highly efficient control for the sealifts system of navy ships connected parallel are realized. Secondly, the marching launch dynamics of controlled self-propelled artillery is presented, and the simulation results are validated by tests. This study provides an important theoretical basis for dynamics design of self-propelled artillery.This paper provides a new highly efficient method for dynamics and control design of controlled mechanical systems. The method in this paper has been used to solve many important engineering problems, and obtained one Chinese national prize in science and technology,6 national invention patents, and 4 national software copyrights.