Russia is actively developing domestic electronics. In this regard, new effective methods of quality control of materials, elements and blocks of electronic equipment (REA) are needed. Scientists at Tomsk State University, with the support of the Russian Science Foundation, have developed a mathematical model and software for flaw detection of images of an intelligent 3D X-ray microtomograph. The approach created at TSU is already being used in industry for the examination of electronics and other equipment.
–Modern electronic equipment contains a huge number of radio components (parts), for example, printed circuit boards, connectors, microcircuits, resistors, which may have external and internal defects," says the project manager, head of the international laboratory "Vision Systems" of TSU Scientific Management Vladimir Syryamkin. – A neural network was trained to diagnose them, using 1,500 reference and 10,000 defective images of materials and elements of the REA.
Along with this, digital counterparts of diagnostic objects were used in the learning process: printed circuit boards, transistors, capacitors, inductors, and more. They were also used in the AI training data library, which increased diagnostic accuracy.
– Now our neural network is able to recognize images of various dimensions and colors. The complex algorithm used in it embodies the properties of the so–called artificial intelligence of the first kind, and is capable of solving the most complex tasks," he added. Vladimir Syryamkin.
At the product testing stage, it was found that the developed algorithmic and software for monitoring and diagnosing materials and elements of REA based on digital X-ray 3D microtomograph images surpasses similar technologies in the USA, China, Taiwan and other countries in accuracy, noise immunity and speed.
The results of the project, carried out with the support of the Russian Science Foundation grant, are already being used in industry for flaw detection of REA elements and other equipment. Software based on neural network technologies can be easily adapted into a product quality management system at enterprises of the military-industrial complex and the civil industry of the Russian Federation. Organizational work is underway to implement the results of TSU projects at Roscosmos enterprises, including the Information Satellite Systems (ISS) named after Academician M.F. Reshetnev.
Please note that this press release is based on materials provided by the company. AK&M Information Agency shall not be held liable for its contents, nor for the legal and other consequences of its publication.
