This paper provides a new goal-oriented (GO) method for reliability analysis of repairable systems with multiple fault modes. First, formulas of operators describing components with multiple fault modes are derived based on Markov process theory. Second, qualitative reliability analysis of such a system is conducted by combining the existing GO method with the Fussell–Vesely method. Third, this new method is applied in reliability analysis of a hydraulic transmission oil supply system. Finally, comparing the study results with fault tree analysis (FTA) and Monte Carlo simulation shows that this new GO method is suitable for reliability analysis of repairable systems with multiple fault modes.

This paper proposes a new systematic reliability analysis method for repairable systems with multifunction modes based on the goal‐oriented (GO) method. First, we create a new function GO operator, a new logical GO operator, and a new auxiliary GO operator, deduce their GO operation formulas, and propose some new rules of the GO operation and an exact algorithm with shared signal of the GO method for such systems. Then, we formulate the analysis process of repairable systems with multifunction modes based on the new GO method. Finally, we apply this new GO methodology to reliability analysis of the control system for a heavy vehicle. To verify the feasibility, advantage, and reasonableness of the new GO methodology, we compare its analysis results with those of fault tree analysis and Monte Carlo simulation. We show that the proposed GO method has clear advantages in system reliability modeling and analysis. All in all, this study not only improves the theory of the GO method and widens its application but also provides a new approach for conducting reliability analysis of complex systems quickly and efficiently.

In order to conduct effective reliability analysis of retracting actuator with multi-state (success state, safety failure state and action failure state), we redefine type-3 operator in goal oriented (GO) method to describe three states of main charge of retracting actuator and improve type-15 operator in GO method to describe the logic relations of multi-state output. The quantitative and qualitative reliability analyses of retracting actuator are made based on GO method in this paper. The system state probability of retracting actuator is obtained through quantitative analysis, and its weakness is found through qualitative analysis. The analysis results show that GO method is effective to improve the reliability of retracting actuator, and this method is also feasible for reliability analysis of other complicated initiating explosive systems.

For a diesel engine, operating conditions have extreme importance in fault detection and diagnosis. Limited to various special circumstances, the multi-factor operating conditions of a diesel engine are difficult to measure, and the demand of automatic condition recognition based on vibration signals is urgent. In this paper, multi-factor operating condition recognition using a one-dimensional (1D) convolutional long short-term network (1D-CLSTM) is proposed. Firstly, a deep neural network framework is proposed based on a 1D convolutional neural network (CNN) and long short-Term network (LSTM). According to the characteristics of vibration signals of a diesel engine, batch normalization is introduced to regulate the input of each convolutional layer by fixing the mean value and variance. Subsequently, adaptive dropout is proposed to improve the model sparsity and prevent overfitting in model training. Moreover, the vibration signals measured under 12 operating conditions were used to verify the performance of the trained 1D-CLSTM classifier. Lastly, the vibration signals measured from another kind of diesel engine were applied to verify the generalizability of the proposed approach. Experimental results show that the proposed method is an effective approach for multi-factor operating condition recognition. In addition, the adaptive dropout can achieve better training performance than the constant dropout ratio. Compared with some state-of-the-art methods, the trained 1D-CLSTM classifier can predict new data with higher generalization accuracy.

The reliability optimization has achieved great concern in recent years. Nowadays, many researchers obtain allocation results which can maximize the system reliability subject to the system budget. In these researches, the effect of system’s functions is always neglected or only considering the single main function of system. In addition, there are also no obvious evidences in results to distinguish the importance level of different units. However, complex systems tend to perform multiple functions. What’s more, the use frequency of each function and the combinations of units to realize different functions are not the same. In addition, the use demand of different functions is decided by different task environments, the demand differentiation of functions has led to the usage of frequency of various functions having different levels about reliability. Therefore, the reliability optimization allocation only considering cost constraint conditions is not accurate and will results in disaccord between the obtained results with actual situation. Focusing on the problem mentioned above, a reliability optimization allocation method that considers cost constraint and importance factor is proposed. In this paper, we consider systems consisting of units characterized by different reliability and importance factors. Such systems are multi-function because they must perform different tasks depending on the combination of units. Different functions may work simultaneously. Firstly, the concept of importance factor is defined to describe the importance of a unit and the required importance factor level of system functions in the task is also given. To deal with the differentiation of system functions, the corresponding bound about importance factor are executed when looking for the optimal solution. Similarly, the cost constraint is also forced. Finally, in order to reduce the randomness of intelligent algorithm, a number of optimization are conducted and a rule is proposed to select the most optimal solution from all the optimal solutions which are obtained in every iterative loop. Example of an integrated transmission device is presented. To begin with, we establish the reliability function of system as the objective optimization function. Then, the restraint of budget and different demands of importance factor of system functions are posed. Furthermore, using a genetic algorithm as the optimization tool, the optimization result can be obtained. Finally, the most optimal solution is selected. The results show that, the method, we proposed is more correct and more approximate than the reality. To verify the advantages and engineering applicability of the new method, the results obtained by the new method are compared with the results obtained under different conditions using basic genetic algorithm, without considering functions and the differentiation of functions, to solve the allocation problem of integrated transmission device, respectively. The reliability optimization allocation method presented in this paper can not only consider the constraint of cost but also can consider the diversities of functions, and thus the optimization results will be more approximate in actual situation. At the same time, this paper can also provide guidance for the similar reliability optimization problem.