To the jubilee of LCTA - MLIT

N. D. Dikusar: "N. N. Govorun was an amazing person..."

We keep presenting a series of memoirs of veterans of MLIT, the 60th anniversary of which will be celebrated this year. Today, the leading researcher of the Laboratory Nikolay DIKUSAR is telling about his life path and colleagues.

In the first year of Faculty of Mechanics and Mathematics of Odessa University, I was presented with the book "The inevitability of a strange world" by Daniil Danin. This is a book about physicists, about Dubna. I read it hard and after a while, I watched the film "Nine days of a year" and I had a desire to come to Dubna. In the fourth year, most students sought to get to practice at the Kyiv Institute of Cybernetics. Our head of computers dissuaded from a trip to Kyiv: they say, there will be thousands of the same and it will not be easy for you to prove yourself. In Dubna, you will find yourself among physicists that respect mathematicians. And so, we, five students, drove for two days in a sitting carriage of the Odessa-Moscow train, then, got to Dubna. The practice went well, N. N. Govorun invited me to do my thesis in Dubna. After defending my diploma in Odessa, a request came from the State Committee for Atomic Energy with my referral to JINR. For some reason, the Ministry of Education of the Ukrainian SSR began to object, I was offered to enter graduate school. A telegram came from Dubna: if you do not send Dikusar to work here, you will lose your place for student practice. I was sent. In Moscow, in the personnel department, they said that there was no way to immediately provide a separate apartment and I already had a family - a wife and a son. They gave a room in a two-room apartment with one neighbor.

I started working at JINR in late January, 1966. When I arrived, Sasha Lukyantsev met me at the station. He said, “You can go back, the Chinese are leaving JINR, there will be a reduction in personnel. At this time, a meeting was held at the Cultural Centre “Mir” to discuss the Institute's immediate prospects. I went to the Centre, found Nikolay Nikolaevich there. He met the issue in five minutes - I was taken to LHEP rate for automation at the JINR Computing Centre that completely moved to the new Laboratory of Computer Engineering and Automation in August, 1966. For the first six months I woke up in a cold sweat at night - I was so used to Odessa and so Dubna was not like it. The job saved me. The lifestyle was very intense. The first impression of the city is quiet, clean, neat, with interesting architecture (the same two- and three-story houses built by captured Germans, I saw later in Sarov), with clean air and good people. We can say - an unusual city. After some time, I began to shake hands with almost every person I met. At that time, you could leave your bike without a lock on the street. Once, it stayed with me for the night outside the store “Dubna” and in the morning, I took it away. Several years passed and the situation changed - my bike was stolen from a locked barn. The supply of the city was also non-standard. Here, I first encountered the concept of "box herring". It was almost completely dried and very tasty. The Volga also attracted, we went kayaking.

Since we began to engage in automation, the first thing N. N. Govorun suggested to me was to develop software for monitoring the quality of measurements and controlling a group of semi-automatic image measuring devices operating simultaneously on the line with the Minsk-2 computer. Female operators independently measured films from track cameras. I first learned about the concept of online that I had to get used to. It was necessary to develop software that monitors the quality of measurements online, with feedback through typewriters. In poor measurement, the typewriter was printed in red that the track should be measured. Qualitative measurements were registered on magnetic tape. Physicists were delighted that the computer controls the quality of measurements by operators. For fun, I added a block that typed poems by Omar Khayyam at the touch of a key. They also liked it, after which I got even more familiar physicists. It was my first job. Electronics from LNP side was developed by A. Shuravin and the communication channel between computers and semi-automatics was developed by V. Shigaev. Then it was pioneering work not only in the USSR, but also in the countries of the socialist camp, N. N. Govorun and V. P. Dzhelepov made reports on this topic at conferences.

N. Dikusar with Chaba Torok

With Yu. Voitenko, O. Zaimidoroga and I. Vasilevsky at HPD

I was also involved in providing mats on BESM-4 to control a scanning machine on a cathode ray tube, on which measurements of images from a spark chamber were taken with the participation of Inna Kukhtina and Valera Zhmyrov but there were no mass measurements on this machine.

The next stage in the development of automation was the task of developing a system of mathematical support for mass measurements and processing images from a five-meter magnetic spark spectrometer (MSS-5). Measurements were carried out on an HPD scanning machine controlled by a CDC-1604-A computer. There was a discussion about who would take on this task. They offered Shigaev but he worked on the task of measuring images from bubble chambers on HPD. They offered someone from physicists but he demanded the organization of a separate sector and a business trip to CERN. As a result, Govorun proposed this task to me with a promise to help people in case of difficulties. I took a chance and agreed. To manage the HPD teams, a prototype of the Yu. I. Shelontsev programme was taken. The processing of data from the MSS-5 was carried out in several stages. At the first stage, frames on film were scanned automatically. Quality control and primary processing of data obtained from HPD were carried out in real time. Geometry filtering and recognition was also done offline on the CDC 6500. The results on magnetic tapes were transmitted to physicists that did further processing according to their programmes. At this time, experiments were implemented on a six-meter spectrometer at the Moscow ITEP under Corresponding Member of the USSR Academy of Sciences V.V. Vladimirsky, one of the designers of the Serpukhov accelerator. We often exchanged processing and recognition programmes. At JINR, A. A. Tyapkin and his colleagues carried out research on MSS-5 as part of the Milan-Bologna-Dubna-Serpukhov collaboration. Some of the data were measured in Dubna, some - in Italy, then, these measurements were combined and processed.

At different stages of work with HPD, the Vietnamese employee Thai Le Thang that had graduated from Belarusian State University, I. Krulyas and M. Shpytko from Poland, I. Amirkhanov, S. Badalyan, M. Haryuzov, A. Raportirenko, Head of Operation Department of HPD V. Moroz and A. Volkov worked with me. With Jan Ruzicka (Czechoslovakia), we developed display software. The employees from the University of Helsinki E. Kuriniemi and T. Hirvonen came on a monthly business trip. During a return trip to Helsinki with V. Inkin, we learnt about the measuring system on the SWEEPNIK scanning machine.

During the operation of HPD, engineers and operators measured with an accuracy of several microns on tracks more than half a million photographs with MSS-5 in an automatic mode. To compare the accuracy of measurements, individual rolls of films were measured both here and in Bologna. Our programme-processed measurements showed a narrower range of errors than those of Italians that used a special processor for primary data processing. After final processing by physicists, a new pseudoscalar meson was discovered and double recharging of negative pions in inclusive reactions on nuclei at 40 GeV/s was studied.

Within the framework of the collaboration, four trips to Italy were held together with A. A. Tyapkin and O. A. Zaimidoroga. I worked in Bologna, at the National Centre for Frame Analysis (CNAF); today, it is the National Centre for Research and Development in Information Technology (INFN) that was headed by Professor M. Masetti. On the first day, I was given a deck of punched cards - the programme code of a special processor. I found one card missing. When I said which card was missing, they took it out of the top drawer of the table and handed it to me - so they checked, a specialist came to them or a spy. In the same building there was a department where an official with a case came from Rome every day. I was forbidden to walk up the stairs of six floors, use the elevator exclusively and say the phrase "io lavoro alla CNAF" to the guard when entering the building. An accompanying employee was assigned to me that always walked with me. I asked him: Antonio, why are you walking with me? He replied: they think you are a spy. A representative of the CIA specially came from Turin, conducted an interview. Anti-Soviet posters were posted on the wall, past which I went to work. Another detail. About two weeks before his departure, Nikolay Nikolaevich Govorun sent me a telegram asking in Latin "when are you returning to Dubna?" The telegram was handed to me with a mysterious smile only on the day of departure.

When Peter I offered the award to Menshikov, he replied: do not need a reward, make me a foreigner. Professor M. Masetti, introducing me to the staff, told me to always give way to the terminal. The next day, when I entered the hall, all the employees, more than 10 people, stood up, giving me their jobs. It was embarrassing, but I felt what it was like to be a foreigner. I want to emphasize that the attitude of colleagues in cooperation both in Milan and in Bologna was friendly, we were invited to visit and treated. When ordinary Italians found out that we were from Moscow, they treated us very well and with great interest.

Nikolay Nikolaevich not only supported me with employees, even gave me his workplace. We then worked virtually around the clock, it was necessary to somehow respond to various failures and Nikolay Nikolayevich's office was not far from CDC 1604-A. It was convenient. We wrote conference papers together. On weekends, we went to pick mushrooms, he often invited me to his house. His mother Maria Antonovna and his wife Raisa Dmitrievna always treated me to something, we had a very good relationship. When he was admitted to a hospital in Moscow, I visited him. Nikolay Nikolaevich took on too much burden: in addition to JINR, he worked in Moscow - at the Academy of Sciences, various committees, in the editorial office of the magazine "Programming" thereby undermining his health. Once, being extremely busy in Dubna, he sent me to Academician A. A. Dorodnitsyn at a meeting of the Committee for the development of computer technology. The talker was an amazing person, very friendly, simple and accessible. They went to his office not one at a time, but with a crowd. He took an informal part in meeting many problems, in particular, with us at HPD. I have never met such people again.

The mathematicians E.P. Zhidkov, G.I. Makarenko, I.N. Silin, V.P. Shirikov, A.A. Korneychuk, A.I. Saltykov, the physicists L.S. Azhgirey, V.I. Moroz, I.M. Salamatin, L.K. Lytkin, the mechanics N.P. Bovin, V. D. Morozov and many others - these are the people I worked next to then. Mikhail Grigorievich Meshcheryakov focused on HPD, as did Rudolf Pose, being deputy director and later, director. As deputy director, he was responsible for HPD and it seems, for the spiral meter. We always greeted him in German when we met. He loved it. Pose supported me during the "perestroika" highlighting a personal computer that was important, at that time, I was dealing with the problem of track recognition.

This problem has grown into the task of finding new effective data processing techniques. Efficiency involves many factors. First of all, this is the accuracy of processing at high speeds, although these are contradictory things: for good accuracy, good programmes are required - very "smart" or very voluminous that entails a loss in speed. It was necessary to look for new ways to meet this problem. Then, I had an idea how to improve techniques for approximating measurements using polynomials in the form of basis elements. Fortunately or unfortunately, the time has come for change and it has become possible to calmly work on this problem. And I have no better place for such work than at JINR. We had complete creative freedom.

The essence of the method of basis elements has a historical trace. In 1853, our great mathematician Pafnutiy Lvovich Chebyshev argued that the Taylor polynomial is poorly suited for meeting practical problems, since it gives high accuracy only in a small neighborhood of one point and with distance from it the error increases greatly. In practice, it is required to make an approximation along an extended segment. It is necessary that the error is uniform throughout the entire segment. He met this problem, the result is the famous Chebyshev polynomials. I used Chebyshev's idea of approximating a smooth function on a segment but I decided differently. Using the internal coupling of an independent variable with three control parameters, a method of basis elements (MBE) based on quadratic and cubic parabolas was developed. Within the framework of this method, MBE polynomials were constructed that provided resistance to errors over a long period, reducing the computational complexity of algorithms and more.

MBE is essential in meeting the problem of polynomial segmentation of 2D curves with a complex topology. This task is relevant in the field of modern information technology, in various fields of scientific research, such as pattern recognition, in digital signal and image processing, in robotics, in the processing of data from physical experiments and more. MBE segmentation of 2D curves of complex topology caused a positive response in research centres in Russia and abroad. MBE is used to meet a number of practical problems both in JINR and in other organizations. MBE polynomials of high degrees are used to improve the methods of processing reactor data in the task of increasing the level of nuclear safety of the IBR-2M reactor. A number of investigations and applications of MBE were co-authored with Csaba Torok (Slovakia). Our cooperation began by chance after one of the seminars back in 1995 and continues to this day. I received great support for MBE research from Heads of the Department George Adam and Jan Bushey.

Of particular interest is the use of MBE for the numerical solution of the Cauchy problem that proposes a fundamentally new method for predicting a solution one step ahead, using a poorly conditioned extrapolation procedure twice. Reviewers found fault with it with the remark that all explicit Adams methods using extrapolation are unstable and unable to meet tough problems. The paradox is that two large extrapolation errors result in a small error in the solution. It was confirmed by a test on a famously tough task. Hard tasks with a sharp transition of processes from one state to another are found in various fields, for example, in chemistry, microbiology, biomedicine. The fifth order of accuracy of the solution is confirmed by comparison with other classical methods and analysis of asymptotically accurate Richardson error estimate.

I would advise young people to turn to the classics more often. In their writings, you can find clues to modern research methods, on the basis of which new results can be achieved. For example, when developing the method of basis elements, the ideas of Taylor, Lagrange, Chebyshev, Gauss, Ponsele and others were used.

And I would wish new creative achievements to the Laboratory in the year of its anniversary and good health, good relations and good luck – to its employees.

Olga TARANTINA,
photos from the archive of N. D. DIKUSAR

Interview in video format