Testing System For Multi-rudders' Performance Based On Virtual Instruments

The system for rudder is the servo mechanism of flight Control System (FCS) in Unmanned Aerial Vehicle (UAV). It receives control signal provided by the computer of FCS, and drives the rudder surface's deflexion. Then the stance and track of flight can be controlled. Therefore the flight Control System's performance and safety are influenced by the system for rudder's capability.The test for rudder system was in analog mode in tradition. This mode had many shortage, such as the test equipment and technology had felled behind, and the automatization level was low, the test parameter was not comprehensive, the test precision was low, and the test result was difficult to synthetic statistic analysis. In order to solve these problems, a new test system for rudder based on virtual instruments should be developed to replace the expensive spectrum analyzer and traditional test system. This new system is more intuitionistic and easily controllable in testing frequency characteristic and stability of the rudder system.In this Thesis, the hardware, software and testing of the system for rudder are introduced. Based on the test requirements of the rudder system in FCS, the design idea of the test system for rudder based on virtual instruments is given. The test method, test case, tests item and playback program for the each parameter of the rudder system are discussed in detail. The real-time quality of the test system is illustrated with emphasis. The scheme for data's real-time storage is given. So the system's real-time performance can be ensured. The system adopts LabVIEW as the software platform. Some functions such as the signals generating, sampling, recurring, data processing and performance analysis are achieved.An efficient method for testing the bandwidth of actuator, linear system identification based on neutral network, is presented in this paper.The paper briefly introduces the working characteristics of actuator,the method for choosing the model of actuator and the identification arithmetic. It also discusses how to chose the input and preprocessing the data for identification.The simulation results indicate that linear neural network based system identification do have very high speed and precision of identification.The interrelated index in different condition of this system has been tested. After analyzing the experimental findings, this system's precision can meet with the requirements demanded. Compared with original manual test system, this test precision, the data preservation, management and analyzing are better.