关键词: 改进细菌觅食优化算法, 风电机组, 异动故障, 故障预警

Abstract: A wind turbine unit abnormal warning technology based on an improved bacterial foraging optimization algorithm was proposed. Based on the overall structure of the wind turbine system, the phased working status of the unit was analyzed, filtering methods were introduced to obtain the benchmark parameters of the unit ’ s operating status, and the characteristic variables of abnormal faults were excavated. Combining with an improved bacterial foraging optimization
algorithm to find the optimal feature subset of faults, and combining with the unit’ s operating data to construct a abnormal warning model, based on this, the computer group predicted the probability of abnormal faults and the threshold of warning indicators, thereby achieving abnormal wind turbine warning. The application results of the example show that the proposed method obtained smaller residual values for early warning and better early warning performance.

Key words: improving bacterial foraging optimization algorithms, wind turbines, abnormal fault, fault warning