Abstract: On account of the complex terrain in the west region of China, the underground mining could break the stability of slope and lead to the slope failure easily. Under the superposed influences of underground mining and landslides pressure, the surface subsidence in valley bottom areas is obviously smaller than that in the flat areas with the similar geological and mining conditions. To accurately predict the surface subsidence in valley bottom areas, a simply supported beam and probability density function (PDF) are employed to update the prediction model. The physical meanings and computing methods of parameters of the updated prediction model are confirmed. The fitness function is built based on the principle that the sum of squares of the difference between the measured value and the predicted value should be the minimum. A new method of parameter inversion is proposed based on the simulated annealing particle swarm optimization (SAPSO), and the corresponding parameters inversion program is developed with MATLAB language. Finally, the research results are applied to a mine in Shanxi province, and the error of mean squares of prediction values is 73 mm which is basically in accordance with the measured values in valley bottom area.

Key words: mining landslide, updated model, simulated annealing particle swarm optimization algorithm, parameter inversion