Expected Effectiveness Evaluation and Optimization Methods of Slope Site Investigation Program

Current evaluation indicators for depicting the expected effectiveness of exploration program often fail to reflect physical process, and the involved parameters are difficult to determine. Besides, the exploration layout strategies in existing optimization frameworks usually rely on the positional relationship, and require determination of the investigation range in advance. To solve the above problems, taking the undrained shear strength parameter as an example, in this paper it proposes and employs the expected reduction rate of the root mean square error (ERRS) to quantify the expected degree of improvement in the concentration effect of factor of safety assessment results towards the reference factor of safety due to the incorporation of reference exploration data, which serves as an indicator to evaluate the expected effectiveness of exploration program. Additionally, a framework for optimizing exploration program is constructed using this indicator in conjunction with a greedy algorithm, aimed at optimizing the locations and number of exploration points. The ERRS indicator calculation process employs the Cholesky decomposition-based midpoint method and the improved Bayesian updating method to discretize both unconditional and conditional random fields. The multiple second-order response surface is used as a surrogate model to replace the deterministic slope stability analysis of spatially variable slopes, significantly enhancing the precision and efficiency of the indicator calculation. Example of undrained saturated clay slope shows: the proposed ERRS indicator can achieve evaluation results close to other indicators without the need to determine complex parameters; the constructed exploration program optimization framework can provide an improved arrangement of exploration points under a specified number of explorations, resulting in more cost-effective and better-performing exploration program without the need to predefine the exploration range. The proposed indicator and optimization framework can serve as a reference for evaluating and optimally designing exploration program in practical slope engineering projects.