ANALYZING MINING TUNNELLING ROCK FAILURE PROBABILITY BASED ON PARTICLE SWARM OPTIMIZATION-SUPPORT VECTOR MACHINE
Since the mining tunnel geological body has characteristics of uncertainty and randomicity, and the conventional certainty calculation has limitations. Moreover, the numerical simulation method expends much time and can not express the relation between parameters and responding value. It brings difficulties in uncertain calculation. Aiming at these problems and general responding surface method limitations, this paper proposed a surrounding rock failure probability calculation method based on particle swarm optimization (PSO)- support vector machine(SVM). According to random distribution of each parameter, adopts orthogonal and uniformity design principles and produces data samples by numerical tests, constructs support vector machine model mapping parameters and responding values which is taking advantage of the nonlinear generalization ability of SVM. The PSO is used to search the parameter of SVM which affecting the forecast accuracy. Afterwards, the PSO-SVM mapping model combined with Monte-C lo sampling method can be used to calculate rock failure probability which can directly use the current large sc lc numerical simulation software. Appling the method to Jinshandian Ore Mine tunnel analysis, the analyzed results show this method is feasible.
JUN XING JING-PING QIU AN-NAN JIANG CHUN-YAN BAO
School of Resources & Civil Engineering, Northeastern University Shenyang, 110004, P.R.China Institute of Highway and Bridge, Dalian Maritime University Dalian, 116026, P.R.China
国际会议
The 7th International Symposium on Rockburst and Seismicity in Mines(2009年第七届国际岩爆与微振动性学术研讨会)
大连
英文
1525-1532
2009-08-21（万方平台首次上网日期，不代表论文的发表时间）