Dynamics Analysis & Motion Control OfAirship

The proposition and rising of the stratosphere platform makes airship attract people'sattention again in resent years. Based on the innovation project of Chinese Academy ofScience, this paper makes a research about the first stage of the project, 1000 cubic meterscertification about the airship in the control system.Dynamics modeling is the base of the airship control problem. In this paper, the 6 DOFequations of motion of the airship are deduced. The Gravity and buoyancy, aerodynamics,thrust suffered by the airship are also analyzed and modeled in the paper. Furthermore, weanalyzed the influence of apparent-mass on the airship and then get the calculation method ofapparent-mass caused by wind.After the dynamics modeling is constructed, the motion equations of the airship arelinearized under the nominal condition with no wind, zero attack angle and sideslip angle, andthe open loop controlled variable can be obtained at the same time. Based on this, the paperanalyzed the stability, dynamic characteristics of the airship, divided the Modal characteristicsof the airship motion and revealed their physical cause of formulation.Based on the linearized equations above, the paper analyzes the uniform circular capacity,uniform cruise capacity, and uniform heave capacity of the airship with different nominalspeed. We can draw from the analysis that the manipulative capability of the airship is closelyrelated to its aviation speed. Thus influence of the speed on the controller must be taken intothe design of the control rules. We here designed a gain scheduling PID controller for thebasic loop of the airship motion, and a feedback gain table based on the airspeed, to achieve agood control on the basic loop of the airship.The uncertainty of the airship model and the particularity of the airship motion bring a lotof problems for the model-based controller. We try to use the neural network method on themotion control, and identified the course heading of the airship using the DTNN neuralnetwork, and then accomplished the objective of heading control by neural PID network.Finally, the paper constructed an airship digital simulation platform and verified theeffectiveness of the control rules by use of the platform.