Recognition of emergency situations of large hydrogenerators and turbogenerators by multi factors analysis of complicated stressed states of units and component parts

• Дата: 2015
• Автори: Вакуленко, А. Н., Кобзарь, К. А., Третьяк, А. В., Гакал, П. Г., Партала, А. А., Овсянникова, Е. А., Морозинский, М. И.
• Формат: Стаття
• Мова: Russian
• Опубліковано: Інститут енергетичних машин і систем ім. А. М. Підгорного Національної академії наук України 2015
• Теми: hydrogenerator; rotor; heat state; damage; complicated stressed state
• Онлайн доступ: https://journals.uran.ua/jme/article/view/57538
• Назва журналу: Energy Technologies & Resource Saving

The analysis of possible causes of emergency situations emerge of Hydrogenerators (Hydrogenerators-Motors) at different operation modes is carried out. Mechanical calculation of complicated stressed state of one of the construction elements of Hydrogenerator-motor is made in details. The values of stress, temperature and movement in the calculated unit namely inter-pole jumper are shown. The method of using of neuron networks to simulate emergency situations of Hydrogenerators, providing search for optimal solutions to prevent failure is grounded. Detailed algorithm of the expert system to ensure trouble-free operation of Hydrogenerators in real time is indicated. The resolve of the task shall be the analysis of changes in specific parameters of generator units in the process of defects emerge and the creation of the knowledge base that stores data about basic design parameters in the origin, development and ultimately damage of the units. It's appropriate to perform analysis using algorithms that can determine the cause-consequence relationships change the basic parameters of the design. Taking in to consideration multi-factors task is the most appropriate algorithm based on the use of neuron networks. Networks shall let determine the service life of the units in real time, based on modes of generator signals transmitted from the temperature detectors, data on vibration condition.