The project “Development of cognitive Smart-technology for intelligent control systems of complex objects based on artificial intelligence approaches” (2018-2020).
The objectives of the project are:
1. Conducting research in the field of bioinformatics for the development of Smart – technologies of creating systems for predicting and controlling complex dynamic nonlinear objects with various types of parameter uncertainties based on various bio-researched approaches of artificial intelligence and, in particular, the development of a promising direction of artificial immune systems.
2. Development of Smart technologies based on the creation of new modified algorithms for artificial immune systems and practical applications using them for technical, technological and social and economic complex control objects in industrial automation systems, technological processes in the oil and gas industry, education and pharmacology.
Since the leading role in the economy of Kazakhstan is assigned to the oil and gas industry and due to the growing requirements for modern industrial enterprises, as well as the rapid development of new information technologies, the development and implementation of effective intelligent control systems and diagnostics of industrial equipment in this industry is relevant.
– Development of an effective Smart-technology for constructing dynamic intelligent control systems for complex objects based on the cognitive approach and the latest AI developments (artificial immune systems, swarm intelligence algorithms, neural networks, genetic algorithms, fuzzy set theory and multi-agent systems) for various applications.
– Synthesis of multifunctional artificial immune system consists of subsystems that implement the basic mechanisms and algorithms of functioning of the human immune system (molecular recognition, clonal selection and negative selection) for assessment and prediction of the behavior of intelligent systems, diagnostics equipment, support decision making and operational adjustment of system behavior.
– Creation of new modified IIS algorithms with the use of swarm intelligence algorithms, neural and genetic algorithms, as well as the development of software for their implementation in multifunctional IIS.
– Development of a diagnostic system for industrial equipment based on the proposed modified IIS algorithms, AMDEC approaches (Analyse Des Modes de Défaillance set des leurs Effect set leur criticité, analysis of operating modes and failures, their impact and degree of criticality) and modern microprocessor technology.
– Creation of mnemonic diagrams to control complex objects using the latest achievements in the field of artificial intelligence and cognitive technologies.
– Implementation of a multifunctional IIS based on a multi-agent approach using cognitive agents in the synthesis of intelligent control systems for various applications in industrial automation systems, technological processes in the oil and gas industry and education.
3. Creation of innovative intelligent information technologies of distance learning (with the possibility of remote access to modern existing industrial equipment of various manufacturers in the laboratories of collective use) using various non-traditional bio-tested approaches: artificial immune systems; neural networks, genetic algorithms, swarm intelligence algorithms, etc.
The problem of training professional engineers to work with modern technologies and complex industrial equipment is solved.
During the implementation of the project for 2018, the following main re-sults were obtained:
– Smart-technology for constructing intelligent control systems for complex objects based on the cognitive approach and the latest artificial intelligence developments has been developed.
– An analytical review of the current state of Artificial Immune Systems (AIS) has been carried out.
– Theoretical foundations have been developed for creating a multifunctional Artificial Immune System.
– The main control mechanisms based on distributed control systems for Honeywell DCS are considered, as well as the principles of transmission and storage of data for use with predicted events using the AIS approach.
– An architecture of an intelligent system based on a modified IIS (clonal selection) algorithm for integration with Honeywell DCS has been proposed.
– A multifunctional AIS has been developed based on modified swarm intelligence algorithms, cognitive technologies and a multi-agent approach.
– A modified particle swarm algorithm with inertia weight (IWPSO) has been developed for the multifunctional AIS.
– A modified cooperative particle swarm algorithm (CPSO) has been developed for the multifunctional AIS.
– The results of modeling modified particle swarm algorithms (IWPSO and CPSO) based on real production data of the Tengiz Chevroil oil and gas company for the diagnostics of industrial equipment (using daily measurements from the sensors of the U300 installation) based on the AIS have been obtained.
– A comparative analysis of the simulation results of the modified particle swarm algorithms (IWPSO and CPSO) with the classical particle swarm algorithm (PSO) has been carried out.
– An innovative cognitive smart technology based on AIS has been developed for distance learning people with visual impairments in engineering specialties in the Honeywell collective-use laboratory for training industrial equipment using the Experion PKS distributed control system. The use of cognitive approach allows to provide high-quality personalized distance learning depending on the type of Central nervous system of students (choleric, melancholic, etc.), psychophysiological features of perception and assimilation of current information, as well as features of vision taking into account the psychotype. In the future, the proposed cognitive Smart technology can be used in the development of mnemonic diagrams for managing complex technical, technological processes and providing information support for human activities, taking into account individual psychophysiological features of perception and awareness of current information.
– A unique immune network technology has been developed for constructing intelligent systems of predicting and managing complex objects under uncertainty of parameters based on the biological approach of artificial immune systems (AIS). This technology for processing and forecasting multidimensional data is aimed at reducing generalization errors and increasing the reliability of the forecast based on the properties of homologous proteins.
The research results are applied in the development of the following applications: intellectualization of industrial automation systems, diagnostics of industrial equipment, with a computer-aided molecular design of drugs with desired properties, distance learning of the engineering specialties of people with visual impairments.
The project “Development of a hardware-medical complex of assessing the psychophysiological parameters of a person” (2018-2020).
Scientific supervisor: Mazakov Talgat Zhakupovich, Dr. Sc. (Phys.-Math.),Prof.
The aim of the project is to develop new methods and technical means of assessing the psychophysiological parameters of a person.
The fundamental objective of the project is the development and research of methods and means of assessing the psychophysiological parameters of a person.
To solve this problem, the following main sub-goals are set:
Scientific novelty lies in the study of existing, as well as in the development of new mathematical models and algorithms to solve the problem of developing criteria for psycho-physiological identification of a person based on interval mathematics. Practical significance consists in the development of methods and software and hardware for obtaining a psychophysiological portrait of a person, which can be applied by government and law enforcement agencies.