During the movements of aircraft, missiles, and ships, the rudder plays an important role in their direction control. In order to test the parameters of the rudders, we have to manually measure each item one by one in traditional production and manufacturing of rudders, which waste a great quantity of manpower and time. In this paper, we present a new application in rudder fault test by using machine learning technology and recommend a new intelligent method for fault location. The main subject revolves around prediction-oriented problems of multi-dimensional performance parameters data mining and the modeling of classification, including the analysis and processing of data features and the solution of fault location based on classification model. In addition, to improve the accuracy of the classification model, we optimized the random forest (RF) algorithm with the shuffled frog leaping algorithm (SFLA), which we call shuffled frog leaping algorithm-based random forest (SFLA-RF). It effectively solves the problem of voting competition among each tree, which makes the decision of the model more efficient and accurate. In a word, by means of automatic test and intelligent analysis, this new method breaks through the technical bottleneck of low efficiency of parameters test and the shortcomings of traditional rudder fault location.

Fiber Optic Gyroscope (FOG) is a new type of rotation sensor that plays an important role in navigation and guidance system. As the sense element of FOG, the fiber coil (FOC) performance directly affects the FOG measurement precision. Many factors influence the quality of FOC, ideal structure and constant small tension control in the winding process were key factors. In order to prevent FOC edge defect, we proposed piecewise variable speed winding method to make regular FOC structure more accessible. In view of the shortage of the common control method used in FOC winding, we optimized fuzzy PID by introducing domain adjusting factors and designed a new tension controller based on mechanical analysis. The optimization controller can adaptively adjust the control factors to ideal settings and ensure constant tension throughout all variable winding speeds stages even in the present of complex external disturbance. Comparative experiments shown that the new controller has a higher accuracy and better robustness than commonly used controller.

Rudders are the important components of aircrafts, missiles and ships, and their traditional test equipment is not intelligent enough, so we have to evaluate their performance by observing every parameter manually. This situation makes it impossible to test the rudders rapidly and in quantity. In this paper, we present a new application in the field of rudder test based on machine learning (ML) and describe new methods for performance evaluation and state prediction. The main topics are concentrated on prediction-oriented problems of multiple performance data mining and modeling: analysis and extraction of data feature, performance scoring based on regression algorithm and cross-validation, screening of defective products and fault location based on classification algorithm and accuracy evaluation. Besides, we propose a new optimized decision tree algorithm (SFLA-MWDT) which solves the common decision difficulty in tree models caused by low-precision decision and high-vote competition. Here, through ‘automatic acquisition + intelligent analysis’, we break through the shortcomings of traditional rudder testing methods and technical bottleneck of low parameter testing efficiency. This test method is applicable to those rudders that have already produced but not yet in use. Also, it provides guidance for the production and practice of rudders.

No abstract