Design Of Weapon Gun Barrel Attitude Angle Measurement System

Modern war is a contest of high technology. Weapons power, as a core performance, has always been regard as the most important in the development of weapon industry. Weapons must be tested first after manufactured and then can be put into battlefield. The testing and evaluation of weapons system should be able to detect various performance parameters of the weapons, base on these test data, the ultimate evaluation results are given. The attitude angle measurement system as a unit of the weapon testing system, requires to measure the real physical rotation angle of the gun barrel. According to the requirements of design, the multi-sensor redundant measurement is advance to overcome various errors in the measurement process, based on profound investigation into the measurement mechanism. A two-layer information fusion model is created to integrate the data detected from the sensors.First of all, the paper takes a deep analysis on traditional angle measure combined with the design requirements. The model of error mechanism was created to achieve compensation of different error factors according to various error sources which exist in the measurement process. After that the paper advances a method of multi-sensor measurement to eliminate the error. At the same time, the author designs the structure of the measurement system.Then, according to general structure of the system, the author designs the hardware, including the design of non-slip friction wheel, the structure of mechanics of the detection, the selection of various sensors and the collection card. After that, the software of the measurement is designed.Furthermore, according to the characteristics of multi-sensor data acquisition, a two-layer of information fusion model is created to satisfy the requirement of data processing. Combining the characteristics of slip signal of friction wheels, the method of correlation analysis is used to identify and repair slip signals and the wavelet analysis as a effective method is used filtering noise and interference in the process of signal collection.Finally, in order to acquire accurate and reliable result from various redundant data, artificial neural network is used to integrate preprocessing signals in the feature level and the outcome of fusion is simulated.