Research On Model And Algorithm For Mission Reliability Allocation Of Spaceflight TT&C System

Mission reliability allocation of TT&C system is a process of dividing the specified mission reliability metric value of system into its subsystems and components according to the corresponding TT&C mission(single task or multi-tasks) and dynamic constrains such as time and space correlation and logical structure of ground stations, tracking ships, relay satellites, control centers routers. It is a kind of combined optimization problem with various types of constrains and complicated relationships among many factors involved. So it has profound application background and value for further research on the model and algorithm, which can provide quantificational support for the general reliability design of TT&C system and put forward the reliability requirements of each subsystem, and make scientific and reasonable TT&C alternatives, as well as ensure the TT&C mission successfully to be completed. The main work and contribution include:(1) Building mission reliability allocation model of TT&C systemAccording to the introduction of composition, structure and characteristics of TT&C system, and the analysis of the basic principle and process of reliability allocation, this dissertation establishes the TT&C system reliability index system which defines the mission reliability allocation index and evaluation index. A standardized description of mission reliability allocation is given from two aspects of TT&C resources and missions and stored in XML format file. Then, the mission reliability description model and allocation model are respectively built based on the analysis of influence factors and constraint conditions, thus providing the modeling basis for solving the reliability allocation problem of TT&C system.(2) Proposing a heuristic algorithm for mission reliability allocation of TT&C systemBased on the related improvement and design of basic heuristic algorithm(HA) and combined with the characteristics of TT&C mission reliability allocation, a heuristic algorithm to mission reliability allocation of TT&C is proposed which is based on the weight information. It can effectively solve the multi-phased mission reliability allocation problems which may have various kinds of logical structure such as series, parallel and backup, and make the predict value of mission reliability allocation to achieve the target value. The iterative direction and velocity controllable strategy are designed to improve the local searching performance and iterative speed of HA and advance the efficiency of the mission reliability allocation. The result of a case study shows that the algorithm has good convergence. By comparing the allocation result errors among different missions and comparing the changes of reliability index of the same resource which participates in different missions, it can be illustrated that the heuristic algorithm has small allocation errors and suitable for solution of different mission structure problems.(3) Designing the particle swarm optimization algorithm for mission reliability allocation of TT&C systemTo solve mission reliability allocation problem in large-scale scenarios, a particle swarm optimization(PSO) to mission reliability allocation of TT&C system is proposed based on the summarization and analysis of the principles and methods of current PSO. Dynamic adjustment strategy of inertia weight coefficient of PSO is designed which effectively reduces the computation of the optimization process and saves the searching time. Initially the inertia weight coefficient is set as large as possible to make it more quickly and more easily to search the global solution and then the inertia weight coefficient will be set as small as possible to search for better local solution. Particle swarm individual velocity controllable strategy which can adaptively adjust velocity direction and scale according to particle characteristics is designed to improve the searching efficiency of the algorithm and avoid getting into local extreme value. Experiment results indicate that the algorithm doesn’t allocate very high MTBF to some resources and can make the predict value of mission reliability allocation more close to the target value than HA. And the running time of the algorithm is better than the HA.(4)Proposing a optimization method of mission reliability allocation based on RBFNNFor PSO algorithms, it requires for a large number of predicted calculation, which may affect the allocation efficiency to finish mission reliability allocation of TT&C system. This dissertation proposes a mission reliability allocation model based on RBFNN. Meanwhile, RBFNN is a kind of meta-model method, which can replace the original system model for operation and has better fitting effect. Adaptive hybrid learning algorithm(AHL) is proposed to avoid easily premature convergence in RBFNN parameter learning process. Principal component analysis is used to determine the initial number of hidden units. Intelligence optimization algorithm combined with decaying gradient descent information is used to training parameters of RBFNN which is improved to accelerate convergence and guaranteed the result to converge at the global optimum with high probability. The algorithm was employed to the offline training of model parameters by sampling the input/output data of the system, and the realization details are provided also. By simulation, it shows that the AHL has higher generalization power and its allocation result is more closed to target value, and the running time of this algorithm is better improved than PSO.