This paper develops a new adaptive interpretable fault detection and diagnosis model (AIFD ) based on belief rule base (BRB) for complex system. The developed AIFD model aims to solve three problems in engineering practice: few fault samples, complex system mechanism and loss of interpretability in modeling process. The first two problems can be addressed by fusing expert knowledge and observation data in the BRB expert system. Moreover, to address the problem of loss of interpretability of AIFD model in its modeling process, a new belief rule weight adjustment method is proposed based on its sensitivity coefficient. The new adjustment method can improve the confidence degree of the users to the FDD output. The developed adjustment method can guarantee the fault diagnosis accuracy and the model interpretability. Then, to further address the influence of uncertain expert knowledge, a new optimization model is developed for the BRB based fault diagnosis model. To illustrate the effectiveness of the developed model, a case study of electric servo mechanism is conducted in this paper.

In order to improve the measuring accuracy of the Hemispherical Resonator Gyro under variable temperature, aiming at the problem of "external temperature is unavailable and internal temperature is unmeasurable," a multiple regression based method is proposed for compensating temperature error in the gyro. The relationship between the internal temperature and the resonant frequency of the gyro is analyzed theoretically. According to a constant temperature experiment, a linear relationship between them is derived based on the least square method. The analysis of a temperature-rising experiment shows that the correlation of the gyro output with the internal temperature is much higher than that with the external temperature. Therefore, taking the resonant frequency as an independent variable, a multiple regression model is established for compensating the temperature error. The compensation effect of the model is verified by temperature-rising and temperature-dropping experiments, which show that the output sequence before compensation is not stable, while it is stable after compensation. After compensation, the drift of the gyro decreases by 62.76% and 48.48%, respectively, and its measuring accuracy becomes equivalent to that at the constant temperature. The experimental results verify the feasibility and effectiveness of the model developed for indirect compensation of temperature error.

The performance of complex systems is closely related to their security and reliability. Accurate performance evaluation is critical for timely system maintenance. Since the system monitoring information obtained usually contains noise, using this information to evaluate the performance of the device may result in inaccurate results. Interval values are used to represent the uncertainty of the monitoring information containing noise in the case of small samples. The uncertainty in the evaluation results is divided into two categories: the uncertainty of the evaluation model and the uncertainty of the monitoring information. The effects of the two categories on the evaluation results and the monotonicity of the output results on the monitoring information and model parameters are quantitatively analyzed. To avoid the problem of interval expansion during model optimization, an IER optimization model based on the belief degrees of monitoring information is built, and the whole evaluation process is solved and optimized uniformly. The monotonicity analysis results simplify the solving process of the IER optimization model, which reduces the difficulty and complexity. Validity and normalization tests are performed on the belief structure of the evaluation results, eliminating the invalid and redundant parts, and obtaining the best interval range of the evaluation results. The effectiveness of the proposed method is verified by its application in an electric servo mechanism.

This study presents a novel dynamic method for evaluating the performance of high-valued complex systems. It aims to address the following 3 problems in performance evaluation of this type of system, including limited observation data of system states, uncertainty in expert knowledge, and high real-time requirement. To address these challenges, we expand the discernment frame of the Belief Rule Base (BRB) to the power set (BRB-P), enhancing its capacity to handle limited data and uncertain expert knowledge. The belief rules are integrated using the evidential reasoning algorithm, ensuring the interpretability of the framework. Simultaneously, we introduce a new optimization model to enhance the real-time performance of the framework and mitigate the explosion of belief rule combinations when dealing with multiple input features. Furthermore, we develop a utility-based framework reduction method to adaptively adjust the discernment framework of BRB-P. This enables the structure and parameters of the framework to be dynamically trained using real-time observation data. An experimental illustration is presented to demonstrate the effectiveness of the proposed framework.

In the scenario of sensor performance degradation, the predictive management of the inertial navigation system needs to maintain reliable accuracy. However, traditional methods are difficult to effectively handle the problem of data fusion of gyroscope with heterogeneous accuracy and dynamic degradation. This paper proposes a new data fusion method for redundant gyroscope systems based on reinforcement learning. This method adopts the soft actor-critic (SAC) reinforcement learning algorithm, breaking through the limitations of traditional models that rely on pre-defined noise models or system models. It achieves adaptive real-time optimization of fusion weights through dynamic interaction with the environment. Based on the principle of information non-increase, a three-stage data fusion model is constructed to precisely match the entire life cycle state of the gyroscope from mild degradation, moderate degradation to severe degradation. To solve the problem of objective determination of stage boundaries, a intersection point detection network is designed. Through convolution feature extraction, bidirectional GRU temporal modeling and multi-branch fusion architecture with difference perception, the critical transition point of sensor stability mutation is automatically identified. Experiments are verified on the inertial instrument control platform. The results show that the accuracy loss in the mild degradation stage is only 7.67%. The fusion accuracy in the severe degradation stage is improved by more than 16% compared to the weighted least squares method (WLS) and the fruit fly optimization weighted least squares method (FFO-WLS), and the single-step inference time is as low as 0.82ms. It has the characteristics of high accuracy, strong adaptability and efficient execution, providing active maintenance support for predictive management.

To address the issue of decreased positioning accuracy caused by interference or blockage of GNSS signals in vehicle navigation systems, this paper proposes a GNSS/MEMS IMU array fusion localization method based on an improved grey prediction model. First, a multi-feature fusion GNSS confidence evaluation algorithm is designed to assess the reliability of GNSS data in real time using indicators such as signal strength, satellite visibility, and solution consistency; second, to overcome the limitations of traditional grey prediction models in processing vehicle complex motion data, two key improvements are proposed: (1) a dynamic background value optimization method based on vehicle motion characteristics, which dynamically adjusts the weight coefficients in the background value construction according to vehicle speed, acceleration, and road curvature, enhancing the model’s sensitivity to changes in vehicle motion state; (2) a residual sequence compensation mechanism, which analyzes the variation patterns of historical residual sequences to accurately correct the prediction results, significantly improving the model’s prediction accuracy in nonlinear motion scenarios; finally, an adaptive fusion framework under normal and denied GNSS conditions is constructed, which directly fuses data when GNSS is reliable, and uses the improved grey model prediction results as virtual measurements for fusion during signal denial. Simulation and vehicle experiments verify that: compared to the traditional GM(1,1) model, the proposed method improves prediction accuracy by 31%, 52%, and 45% in straight, turning, and acceleration scenarios, respectively; in a 30-s GNSS denial scenario, the accuracy is improved by over 79% compared to pure INS methods.