Abstract:How to accurately quantify the uncertainty of stochastic dynamical responses affected by uncertain loads and structural parameters is an important issue in structural safety and reliability analysis. In this paper, the conditional uncertainty quantification analysis for the dynamical response of stochastic structures considering the measurement data with random error is studied in depth. A method for extracting the key measurement condition, which holds the most reference value for the uncertainty quantification of response, from the measurement data is proposed. Considering the key measurement condition and employing the principle of probability conservation and conditional probability theory, the quotient-form expressions for the conditional mean, conditional variance, and conditional probability density function of the stochastic structural dynamical response are derived and are referred to as the key conditional quotients (KCQ). A numerical method combining the non-equal weighted generalized Monte Carlo method, Dirac function smoothing technique, and online-offline coupled computational strategy is developed for calculating KCQs. Three linear/nonlinear stochastic dynamical examples are used to verify that the proposed KCQ method can efficiently and accurately quantify the uncertainty of the structural response considering measurement conditions. The examples also compare the traditional non-conditional uncertainty quantification results with the conditional uncertainty quantification results given by KCQs, indicating that considering measurement conditions can significantly reduce the uncertainty of the stochastic dynamical responses, providing a more refined statistical basis for structural safety and reliability analysis.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is: how to accurately and efficiently quantify the uncertainty of random linear/non - linear structural dynamic responses when considering measurement conditions with random errors. This problem specifically involves estimating conditional statistical characteristics such as the conditional probability density function (PDF), conditional mean, and conditional variance of the response under given conditions, that is, the conditional uncertainty quantification problem.
### Background and Motivation
In practical engineering, the dynamic responses of structures are often affected by various uncertain factors, for example:
- **Uncertainty of geometric parameters**: Due to errors in the assembly process.
- **Uncertainty of material parameters**: Due to fluctuations in the material preparation process.
- **Uncertainty of external loads**: Such as random loads caused by natural phenomena like earthquakes, winds, and sea waves.
These uncertainties make traditional deterministic algorithms no longer applicable, and it is necessary to develop effective uncertainty quantification methods to evaluate the influence of these factors on structural responses. In addition, in many engineering problems, engineers usually arrange measurement points on the structure to obtain structural response data. If the uncertainty can be quantified by combining these measurement data with random errors, the uncertainty assessment of random responses can be further refined, providing a more reliable basis for the safety and reliability analysis of the structure.
### Limitations of Traditional Methods
Most of the existing uncertainty quantification methods are unconditional, that is, only the uncertainties in the structural model and load are considered. This may lead to the calculated uncertainty of the random structural dynamic response being too conservative and having limited reference value. For example, when using the three - standard - deviation interval to measure the uncertainty of the response, if the range of this interval is large, it has limited significance for understanding the dynamic behavior of the structure.
### Research Objectives
The main purpose of this paper is to study how to accurately and efficiently quantify the uncertainty of random linear/non - linear structural dynamic responses when considering measurement conditions with random errors. Specifically, it includes the following points:
1. **Propose the Key Conditional Quotient (KCQ) theory**: By extracting key conditions from the measurement data and using the principle of probability conservation and conditional probability theory, derive the exact quotient - form expressions of the conditional PDF, conditional mean, and conditional variance of the random structural dynamic response.
2. **Develop an efficient numerical calculation method**: Combine the non - equally - weighted Generalized Quasi - Monte Carlo (GQMC) method, Dirac function smoothing technique, and on - line - off - line coupling calculation strategy to calculate KCQs.
### Main Contributions
1. **Proposed the KCQ theory**: By extracting key conditions from the measurement data, constructed the exact quotient - form expressions of the conditional PDF, conditional mean, and conditional variance.
2. **Developed an efficient numerical method**: Use the GQMC method to calculate high - dimensional integrals, smooth the Dirac function involved in KCQ - PDF, and at the same time adopt the off - line - on - line coupling calculation strategy to improve the calculation efficiency.
Through these methods, this paper shows how to significantly reduce the uncertainty of random dynamic responses when considering measurement conditions, providing a more refined statistical basis for the safety and reliability analysis of the structure.