CAJ | 학술논문

电力系统实际运行中常常发生 n-k(k重同时故障)事故，严重时可引发连锁故障。针对目前的随机搜索算法在识别最小n-k事故时计算量大，而现有故障筛选方法大多数只针对导致系统失稳的故障的问题，可利用随机化学算法来有效识别最小n-k事故集合，同时获得大量可能引发连锁故障的最小n-k事故集合。与其他搜索算法相比，寻找一个最小n-k事故集合最多只需O(log(n))次仿真。以中国东部某电网为例，仿真试验运用改进的OPA (ORNL—PSerc-Alaska)模型，验证了随机化学算法识别最小n-k事故集合的有效性。此外，随机化学算法可为识别系统脆性线路提供新的思路。
전력계통실제운행중상상발생 n-k(k중동시고장)사고，엄중시가인발련쇄고장。침대목전적수궤수색산법재식별최소n-k사고시계산량대，이현유고장사선방법대다수지침대도치계통실은적고장적문제，가이용수궤화학산법래유효식별최소n-k사고집합，동시획득대량가능인발련쇄고장적최소n-k사고집합。여기타수색산법상비，심조일개최소n-k사고집합최다지수O(log(n))차방진。이중국동부모전망위례，방진시험운용개진적OPA (ORNL—PSerc-Alaska)모형，험증료수궤화학산법식별최소n-k사고집합적유효성。차외，수궤화학산법가위식별계통취성선로제공신적사로。
In the practical power system,n?k contingencies (k multiple simultaneous failures) do occur, which may trigger sequences of cascading outages when severe. However, the current random searching algorithms (RSA) generally require large computation when identifyingn?k minimal contingencies. In addition, the present contingency screening methods only target the failures which result in system instability. Under this circumstance, a random chemistry (RC) algorithm for identifyingn?k minimal contingencies is described, with massive collections ofn?k minimal contingencies that are likely to trigger cascading outages being obtained. This method requires onlyO(log(n)) simulations per contingency identified, which is orders of magnitude faster than the other RSA. The algorithm is applied to a certain power network in Eastern China with an improved OPA (ORNL—PSerc-Alaska) model, and the simulation results prove the validity of RC algorithm in identifyingn?kminimal contingencies. What’s worth mentioning is that this RC algorithm may offer a new idea for identifying system vulnerability.