The 3rd International Conference on Robotics, Artificial intelligence and Intelligent Control(RAIIC 2024)

Invited Speakers

Session 1: Artificial Intelligence Technology and Applications

Aleksei Shinkevich.png

Professor Aleksei Shinkevich

Academician of the Russian Academy of Natural Sciences,Kazan National Research Technological University, Russia

Shinkevich Aleksei graduated with honors in 1999 from the Socio-Economic Faculty of Kazan State Technological University with a degree in Economics and Enterprise Management (by industry).In 2005, at the age of 28, he defended his doctoral dissertation in two specialties: Economics and management of the national economy (innovation and investment management), Mathematical modeling, numerical methods and software packages on the topic "Improving the institutional system of innovative development of the region (on the example of the Republic of Tatarstan)".


The use of artificial intelligence in the process of predictive analysis and organization of resource-efficient chemical industries


The formation and replication of a methodology for building (including the development of tools, principles, methods, technologies and models) a predictive analytics system for chemical industries can increase the level of technological independence of the Russian industry in terms of building research potential, transferring it into engineering solutions and scaling these solutions within the boundaries of both individual sectors of industry and the economy as a whole. Production systems and industrial complexes have a complex structural and functional organization in the form of a variety of actions, stages, stages, conditions, external and internal connections, which requires integrated engineering and economic management based on mathematical algorithms and models. Forecasting the provision of the production system with technologies and material and technical resources in the context of the development of technological sovereignty is an interdisciplinary large-scale task that goes beyond production and is solved at the level of the production infrastructure of industries and industrial complexes.  

Vladimir Soloviev.pngProfessor Vladimir Soloviev

Financial University,Russia

Vladimir Soloviev is a Professor in the Department of Artificial Intelligence at the Financial University under the Government of the Russian Federation. He holds a D.Sc. and PhD in Mathematical and Instrumental Methods for Economics and is a renowned Russian expert in applied machine learning, data science, and intelligent robotics.

Since 1997, he has been involved in the development and implementation of machine learning systems. From 2016, he has focused on technologies for natural language analysis, processing, computer vision, and video analytics. Additionally, starting in 2020, he has been working on projects related to intelligent robotics in industry and agriculture, where he has developed several unique technologies.

Vladimir Soloviev has been teaching artificial intelligence at Russian universities since 1997 and has been working at the Financial University since 2011. He served as a professor in the Department of Applied Mathematics from 2011 to 2016 and then headed the Department of Data Analysis and Machine Learning until 2022. Currently, he is a professor in the Department of Artificial Intelligence.


Gradient Boosted Decision Trees with Ternary Division by Features


The paper explores the efficiency of ternary and binary division by features in gradient boosted decision trees is compared. The study proposes incorporating ternary logic into gradient boosting, modifying XGBoost to branch into three subtrees instead of two. Experiments on the Breast Cancer and California Housing datasets demonstrate that ternary logic can improve prediction quality, particularly with larger datasets, as evidenced by slightly better LogLoss for Breast Cancer classification and more significant MSE, MAPE, R2 improvement in California Housing regression.


Oleg Kuzenkov.png

Professor Oleg Kuzenkov

Lobachevsky State University of Nizhniy Novgorod,Russia

Oleg Kuzenkov is Professor at the Institute of Information Technologies, Mathematics and Mechanics, Department of Differential Equations, Mathematical and Numerical Analysis, Nizhny Novgorod State University. N.I. Lobachevsky, Nizhny Novgorod, Russia. Oleg Kuzenkov's original achievements cover a wide range of topics in the field of optimization and optimal control, as well as intelligent systems with applications to biological modeling, including aquatic ecosystems and infection spread systems.


Modeling zooplankton behavior using neural networks


The phenomenon of zooplankton daily vertical migrations of was discovered two hundred years ago; it is of the great importance for aquatic ecosystems. The main objective of the work is to find the connection between the local current state of environment and the elementary reactions of zooplankton corresponding to fitness maximum. To solve the problem, a four-layer convolutional neural network is built. For the convenience, we use discretization of daily time and dive depth. The input information for the neural network is the amount of food (phytoplankton), predator (fish), hydrogen sulfide concentration and temperature at the current location of the zooplankton, as well as the degree of illumination at a given time of day. The output information is a signal for the organism reaction during the elementary period of time: an upward displacement by an elementary distance, a downward displacement, or maintaining the previous position. The network is trained without a teacher, aimed at maximizing the corresponding value of the fitness function. The trained neural network implements behavior strategy consistent whit field observations. It has the ability to adapt to changing environmental factors; it forms quasi-optimal behavior in changing conditions without retraining.  Thus, the neural network imitates the behavior of the living organism. The discovered patterns of behavior are important for the study and exploitation of aquatic ecosystems.

Svetlana Novikova.png

Professor Svetlana Novikova

Kazan National Research Technical University named after A.N. Tupolev — KAI (KNRTU-KAI), Russia

Svetlana Novikova is a Professor of Applied Mathematics and Computer Science department, Kazan National Research Technical University, Kazan, Russia. Research interests currently affects decision support systems with artificial intelligence, including artificial neural networks, fuzzy systems, neuro-fuzzy network and others, as well as optimization of such systems. The obtained scientific results are applied in practice, primarily in technical systems of optimal control. Honored Worker of the Higher School of the Republic of Tatarstan. Winner of the Prize of the Academy of Sciences of the Republic of Tatarstan named after Viktor Alekseevich Popov. Winner of the competition "50 innovative ideas for the Republic of Tatarstan". Nomination "Socially Significant Innovations". Project "Software module for monitoring and control of artificial pancreas". Received 22 grants of Russian and international levels. Has 255 publications, including 11 monographs. Has more than 20 patents and certificates of  State registration of the Russian Federation of computer programs.  She worked as a visiting professor in Germany, Finland, France and China.


Intelligent personalized neural network glycaemia forecasting  in diabetic patients based on mixed time series with its prospective implementation into the robotic smart insulin pump (RSIP) system


The authors explore the prospects of implementing a self-learning neural network algorithm within the hardware and software of the intelligent insulin pump (SIP) to personalize insulin therapy. SIP is a self-contained automated system controlling insulin delivery/infusion. It imitates insulin secretion by the pancreas of an average adult. The required dose is calculated based on biometric data provided by the on-body sensors. The dose calculation self-learning neural network algorithm allows personalized treatment, adjusting to the dynamically changing bio-metric values/data in real time. Several possible neural network paradigms were applied and analyzed with a view to forecasting the blood glucose values in short-term (3 minutes) and mid-term (30 minutes) periods using mixed time series data on blood glucose values, active insulin, and carbohydrate values taken at 3-minute intervals. The authors tested the neural network model, and the results of the experiments involving two patients with different insulin indexes are provided. The claim that the standardization of the model for patients with different insulin indexes will fail was underpinned, allowing us to conclude that personalization of the diabetic patients’ treatment is needed. Recommendations are suggested for designing and neural network training to forecast blood glucose values further. Prospects for future insights are outlined.


Associate Professor Jin Xin

Yunnan University, China

Xin Jin received his BS degree in electronics and information engineering from the Henan Normal University, Xinxiang, China, in 2013, and his PhD in Communication and Information Systems from Yunnan University, Kunming, China, in 2018. He was a Post-Doctoral Fellow with the School of Software, Yunnan University from 2018 to 2020. He has published over 100 articles in IEEE Transactions on Cybernetics, IEEE Transactions on Instrumentation & Measurement, IEEE Transactions on Geoscience and Remote Sensing, IEEE Transactions on Consumer Electronics, IEEE Transactions on Emerging Topics in Computational Intelligence, IEEE Transactions on Intelligent Vehicles , Engineering Applications of Artificial Intelligence, Information Sciences, Expert Systems with Application and so on. He has more than 1400 citations, and his H-Index is 19. He has won the prize of Natural Science Award of Yunnan Province Science and Technology Award, the prize of Excellent Achievement Award of Yunnan Province Postdoctoral Research Fund, and the Excellent Doctoral Thesis of Yunnan Province. He served as the Guest editor of Frontiers in Neurorobotics, Life, Discover Applied Sciences, and Intelligence & Robotics.


Image Processing Applications of Generative Models with their Security Research


Image generative models, such as Generative Adversarial Network (GAN) and Variational autoencoder (VAE), have profoundly changed the field of image processing. They are capable of generating realistic images and have shown great potential in tasks such as image synthesis, editing, and enhancement. Continuous innovation and optimization of neural network structures, such as such as GAN and Transformer, have led to remarkable achievements in image processing tasks. This report introduces the applications of image generation technologies in image colorization and fusion. Due to the inherent learning mechanism of deep learning models, they may be misled or even manipulated. Thus, image generative models face security issues, and themselves may generate security generation threats as well, particularly, the issues such as adversarial attacks against models and forged image. This report discuses the adversarial attack for cross-task multi-model system based on the research of adversarial attack and defense for single-task model, and the detection technology for generated images is also discussed.


Associate Professor Xinqiang Chen

Shanghai Maritime University, China

Xinqiang Chen serves as associate professor and Ph.D supervisor at Shanghai Maritime University. He mainly focuses on advanced AI technique-based smart ships, autonomous port management, and computer vision. He has published over 70 papers in top journals and conferences in the fields of transportation, navigation, and logistics. including around 60 SCI papers. He has authored 10 ESI/hot papers. He has authorized and applied for 10 invention patents (e.g., U.S. patent, Canada Patent). He served as PI/Co-PI over 10 National Natural Science Foundation of China/Postdoctoral Fellowships, and 2 university-level fellowships. He has been invited to serve as a reviewer for top SCI journals and conferences in the fields of computer vision, transportation, and smart shipping. He serves as Associate editor for IET Intelligent Transport Systems (SCI, JCR Q3), Measurement and Control (SCI, JCR Q3), International Journal of Advanced Robotic Systems (SCI), Discover Applied Sciences (EI). He served as editorial board member for Journal of Marine Science and Engineering (SCI, JCR Q2) and Sustainability (SCI, Q1).


Visual Navigation and Traffic Situation Awareness for Smart Shipping


Smart shipping serves as a key focus of transportation, maritime, and manufacturing powers, has become a research focus in the field by breaking through the bottleneck of driving environment perception for intelligent vehicles/ports/ships based on methods such as computer vision, artificial intelligence, and deep learning. This seminar mainly focuses on the extraction and analysis of human, vehicle, and ship trajectories in the scenarios of automated docks and intelligent ships. Artificial intelligence, deep learning, computer vision, and other methods are used to carry out vehicle/ship/person detection, tracking, and recognition in low visibility, high interference, and complex shipping environments, and to discriminate and analyze the motion behavior of related objects.

Elmira Kremleva.png

Associate Professor Elmira Kremleva

Kazan National Research Technical University named after A.N. Tupolev — KAI (KNRTU-KAI), Russia

Dr. Elmira Kremleva is  Associate Professor of  Applied Mathematics and Computer Science, Kazan National Research Technical University, Kazan, Russia. Her research interests include artificial intelligence (Fuzzy Logic, Neural Networks, Expert Systems) in control of systems for various purposes, mathematical neural network methods to ensure environmental safety, computer modeling of processes and systems for various purposes, E-learning tools for teaching mathematics including smart systems. Elmira Kremleva is the author of more than 50 scientific publications and has several patents in the field of neural network modeling and intelligent control. Kremleva took part in such international projects as “Development of distance learning intellectual courses in the electronic environment MathBridge” (SAARBRÜCKEN, GERMANY), “Digitalization of the economy of the Republic of Tatarstan through the introduction of intelligent control systems: aircraft construction, medicine, legal proceedings” (Istanbul, Turkey) , "Development of a mass open e-course "Computer simulation of systems" (MOSCOW, RUSSIA), and others.


Automatic typing of aerospace imagery textures using cascaded neural network clustering


The article considered the problem of recognizing objects in an aerospace image when the image was taken from a vast distance (from space orbit), noisy, and distorted. At the same time, we classified image areas, during which each image pixel is associated with the category of terrain objects it represents. In this case, the classes may not be known in advance. Let us name objects whose classes were not specified untyped. Therefore, the classification task for untyped objects will include two subtasks: preliminary class definition (in the future, a new unexplored object will be assigned to one of these); directly classification. This issue will distribute the new object to one of the classes defined in the previous step. We propose a step-by-step method for finding the class of a multidimensional object when the set of classes is not known in advance. The developed method first solves the problem of choosing classes from an untyped heterogeneous set of objects and then classifies an arbitrary new object into specific classes. Classes underlie the author's cascade neural filtering algorithm, and objects are classified using the author's model based on a finite automaton.

Natalia Valitova.png

Associate Professor Natalia Valitova

Kazan National Research Technical University named after A.N. Tupolev — KAI (KNRTU-KAI), Russia

Dr. Natalia Valitova is Associate Professor of  Applied Mathematics and Computer Science, Kazan National Research Technical University, Kazan, Russia.  The main scientific direction is the use of artificial intelligence algorithms in aircraft construction. She is the author of more than 50 scientific publications and has several patents in the field of  intelligent design. Dr. Valitova took part in such international projects as “Development of distance learning intellectual courses in the electronic environment MathBridge” (Saarbrücken, Germany), “Digitalization of the economy of the Republic of Tatarstan through the introduction of intelligent control systems: aircraft construction, medicine, legal proceedings” (Istanbul, Turkey) , "Development of a mass open e-course "Computer simulation of systems" (Moscow, Russia), and others. Was a guest lecturer in the Program of Academic Mobility of Teachers within the framework of the ERASMUS + project (Bari Polytechnic University, Bari, Italy,). In 2023 she was awarded a Certificate of Honor from the Minister of Digital Development of Public Administration, Information Technologies and Communications of the Republic of Tatarstan. 


Hybrid model for predicting an unknown process based on a cluster version of the K-nearest neighbors method


The authors have introduced a new model for predicting the values of an unknown process by combining clustering and k-nearest neighbor methods. One of the key features of this method is that it considers groups of similar vectors rather than individual vectors when determining the nearest neighbors of an unknown vector. When calculating distances between a vector and groups, a pairwise average is used to ensure that all available data is taken into account. Additionally, the method allows for the prediction of an unknown value using a separate model for each group. In the simplest scenario, the weighted average of known parameters within the closest cluster to the vector is proposed, with the weights being inversely proportional to the distance.

Session 2: Robotics and Applications


Professor Yong Song

Shandong University, China 

Prof. Yong Song received the B. Sc. in control science from Shandong University in 2001, the M. Sc. degree in control theory and control engineering from Shandong University in 2008, and the Ph.D. degree in pattern recognition and intelligent systems from Shandong University in 2012. Currently he is a Professor of Shandong University. Professor Song’s research expertise is in the area of intelligent systems control, swarm intelligence & multi-robot control, embedded system technology and applications. Prof. Song serves as the associate editor of the international journal of Intelligence & Robotics. From 2017 to 2018, he was a visiting scholar at the Laboratory of Advanced Robotics and Intelligent Systems, University of Guelph. He is the executive director of the Shandong Provincial Society of Automation, vice chairman of the Weihai Society of Mechanical and Electrical Automation.


Robot Skills Learning Based on Reinforcement Learning and Large Pre-trained Models


Research on robotic manipulation skill acquisition has made significant progress in both modeling techniques and engineering applications in various fields. In this talk, I will start with a very brief introduction to some robot skills learning methods, such as deep reinforcement learning and large pre-trained models. After that, I will focus on our current research on robotic manipulation skill acquisition, including goal-conditioned reinforcement learning with adaptive intrinsic curiosity and universal value network fitting for robotic manipulation, hierarchical reinforcement learning with curriculum demonstrations and goal-guided policies for sequential robotic manipulation, dual-critic deep reinforcement learning for push-grasping synergy in cluttered environment, learning from one-shot visual demonstration in the dense cluttered environment via context translation, and robot task reasoning method based on pre-trained large model.

Mirzoian Mariam.pngDr. Mirzoian Mariam

Deputy Dean of the Faculty of Information Technology and Big Data Analytics,Financial University, Russia

Born in Hrazdan (Armenia) in 1992, I obtained my Bachelor's degree from the Armenian State University of Economics in 2014. I was awarded a scholarship from Rossotrudnichestvo in 2016 to pursue a Master's degree in Business Informatics at the Financial University under the Government of the Russian Federation. Additionally, I received a scholarship from the Ministry of Science and Higher Education of the Russian Federation in 2016 for postgraduate studies. In 2019, I completed my postgraduate studies in Mathematical, Statistical, and Instrumental Methods of Economics at the Financial University under the Russian Government. In 2024, I successfully defended my PhD thesis on Intelligent Decision Support Systems. Currently, I serve as a Senior Lecturer in the Department of Business Informatics at the Faculty of Information Technologies and Big Data Analysis and hold the position of Deputy Dean for Supplementary Professional Education and International Communication at the Faculty of Information Technologies and Big Data Analysis. I am also a member of the Council of Young Scientists at the Financial University.


Implementation of the procedure for searching for personnel from external sources using neural networks


The article discusses how neural network technologies are utilized for the recruitment process in the state civil service, including analyzing numerous resumes to identify the most qualified candidates. It also explores the implementation of intelligent technologies in discovering professional development opportunities for current civil service employees to support their continuous growth and career advancement.

Nikita Andriyanov.png

Associate Professor Nikita Andriyanov

Financial University, Russia

Born in 1990 in Ulyanovsk (USSR), which is also the birthplace of Vladimir Lenin. Graduated from Ulyanovsk State Technical University, Bachelor's degree in 2011 and Master's degree in 2013. In 2017, he defended his PhD thesis on the subject of image processing. In 2018 he was awarded the medal of the Center for Pattern Recognition and Machine Intelligence (CENPARMI, Montreal, Canada). Silver medalist of the 1st World Championship in Computer Science (Dubai, UAE). In 2023, I was an invited speaker at the AI Journey conference. I am actively engaged in research in the field of deep learning and computer vision, I work as an associate professor of the Department of Artificial Intelligence at the Financial University (Moscow) and head of the Youth Laboratory of Computer Vision. 


Development of Intelligent System for a Robot Assistant Based on the Different Modalities Data Processing


The paper presents a microservice architecture to realize intelligence in a robot assistant. In addition, the hardware of the robot was made on a 3D printer, and cameras, sound sensors (microphone and speakers), and a transmitter based on Wi-Fi technology are integrated into the hardware. This allows the robot to work with different types of data. In particular, the presented assistant can recognize a person by face, can understand his speech and synthesize both text and voice response. Large language models are used for text generation, which reduces the criticality of speech recognition errors. Thus, a comprehensive artificial intelligence system is developed to provide convenient interaction between robot (machine) and human. It is also important to note that the developed system can use the API with the class schedule of the Financial University and answer the user's questions about the schedule.

Jiyu Cheng.png

Associate Professor Jiyu Cheng

Shandong University, China

Jiyu Cheng received the B.E. degree in automation from Shandong University, Jinan, China, in 2015 and the Ph.D. degree in department of electronic engineering from the Chinese University of Hong Kong, Hong Kong, in 2019. He is currently an Associate Professor with the Department of Control Science and Engineering, Shandong University. His current research interests include multirobot system, deep learning, and reinforcement learning.


Multirobot Autonomous Collaboration in Complex Scenarios


Multirobot collaboration plays a vital role in many real-life applications, such as inspection, search and rescue and so on. As a main research branch in robotics, it has attracted wide attention and developed rapidly. However, efficient collaboration in especially large and unstructured environments is still a challenging problem. In this talk, we will introduce our recent research on several typical multirobot tasks and talk about our exploration on how the data driven approach can empower the collaboration in a multirobot system. 

Tatiana Gainutdinova.jpg

Assistant Professor Tatiana Gainutdinova

Kazan National Research Technical University named after A.N. Tupolev — KAI (KNRTU-KAI), Russia

Tatiana Gainutdinova is assistant professor of Department of Aircraft Design, Kazan National Research Technical University, Kazan, Russia. Research interests currently concern decision support systems using artificial intelligence, including the development of mathematical models for controlling a flock of autonomous vehicles, automatic landing using computer vision. The obtained scientific results are applied in practice, primarily in technical optimal control systems. Honorary worker of higher education. Has more than 60 publications. He has more than 15 patents and certificates of state registration of the Russian Federation for computer programs.


Modified Flocking Model for  Autonomous Vehicles


The developers of unmanned aerial vehicles have realized the potential utility of mathematical models in the management of artificial systems, for example, a flock of autonomous flying robots that reproduce bird’s flock flight patterns. The simplest models of the flock describe the alignment rule as an obvious mathematical axiom: each object aligns its velocity vector with respect to the average velocity vector of objects in its neighborhood]. To this rule is added the connection of accelerations, preferred directions, and adaptive decision-making schemes for increasing stability at high speeds. In other models, the alignment rule is a consequence of the interaction forces or velocities based on damped dynamics. Studies of smooth collective movement of the flock also include an analysis of the stability of the system by reaction time and possible delay in communication (information transfer), the presence or absence of noise from sensory signals, or unpredictable environmental disturbances, such as gusts of wind. It is known, for example, that if there is a time delay during the exchange of data between flock’s agents system instability may occur. The modified model of the flock formation includes the weighting factors represented by the function of values (the distance between the objects of the flock) allows us to configure the model parameters to obtain different forms of collective movement of autonomous vehicles. The model allows to take into account the influence of selective (or accidental) temporary communication loss between flock’s unit due to certain limit distance between objects is exceeded.  As a unit of the flock, a multikopters (quadcopter) are taken. The flight of autonomous robots is controlled by low-level algorithms based on PID speed controllers, i.e. this low-level control algorithm for an individual object as applied to this case has at its input some desired flight velocity vector. The general movement develop in accordance with the basic flocking principles. This implies a ban on the central processing of group dynamics by an external computer, but allows the use of on-board devices, such as GPS, for external position determination. Simulations have shown that the flock's  modified model with variable weighting factors  depended on the distance of surrounding objects to an arbitrary i-th object can be configured to obtain different forms of collective movement of independent vehicles, up to directional. By the weight coefficient it is possible to simulate situations of information loss due to a considerable distance location of retired objects or its accidental loss. 

wang jiankun.jpg

Assistant Professor Jiankun Wang

Southern University of Science and Technology, China

Jiankun Wang (Senior Member, IEEE) received the B.E. degree in Automation from Shandong University, Jinan, China, in 2015, and the Ph.D. degree from the Department of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong, in 2019.

He is currently an Assistant Professor with the Department of Electronic and Electrical Engineering, Southern University of Science and Technology, Shenzhen, China. During his Ph.D. study, he spent six months at Stanford University, CA, USA, as a Visiting Student Scholar, supervised by Prof. Oussama Khatib. His current research interests include robot motion and path planning, human–robot interaction, and artificial intelligence.

He serves as the Associate Editor of IEEE Transactions on Automation Science and Engineering, IEEE Transactions on Intelligent Vehicles, IEEE Robotics and Automation Letters, International Journal of Robotics and Automation, Biomimetic Intelligence and Robotics, Intelligence & Robotics, and member of Early Career Editorial Board of IEEE/CAA Journal of Automatica Sinica. He also serves as the Program Chair of 2023 International Conference on Biomimetic Intelligence and Robotics, and member of Senior Program Committee of IEEE ICRA 2021, IEEE ROBIO 2019, and IEEE ICIA.


Sampling-based Robot Path Planning and Its Applications


Sampling-based path planning algorithms have achieved great success in robotics due to their ability to efficiently search the state space. In this talk, basic sampling-based algorithms and their strengths and weaknesses are introduced at first. Next, the adaptive sampling strategies and real-time optimization techniques are presented to overcome existing limitations of sampling-based algorithms. Then, AI-driven path planning algorithms are introduced, which lead to more efficient sampling strategies. Finally, a series of specific applications demonstrate the effectiveness and efficiency of the sampling-based path planning algorithms.


Assistant Professor Roman Zimov

Moscow Polytechnic University, Russia

Born on September 21st, 1994, in Lyubertsy, he attended Moscow Polytechnic University from 2012 to 2017, where he studied ground vehicles. In September 2015, he joined the FDR Moscow team.During 2015 and 2016, he worked as an engineer on the development of the rear suspension for a sports prototype. In 2017, he graduated from Moscow Polytechnic University. From 2018 to 2021, he was the head of the chassis department at FDR Moscow, working on the sports prototype Fenix.From 2022 until 2023, he was a racing engineer with the G-Drive Racing team. Since the end of 2022, he has served as the chief engineer and racing engineer for the FDR12 hybrid sports prototype project.


The development of a hybrid sports prototype.


— Hybrid power plants in all the most prestigious and technologically advanced racing series (F1, WEC, WRC).

— Mass versus power.

— Regenerative braking, the ability to collect energy during braking.

— The strategy of using the hybrid part of the power plant

— Who is faster: a hybrid car or a vehicle with a traditional internal combustion engine?

Session 3: Intelligent Control Systems and Optimization


Professor Mingzhe Hou

Harbin Institute of Technology, China

Prof. Mingzhe Hou received his Ph. D. degree in control science and engineering from Harbin Institute of Technology in 2011. He became a Lecturer, an Associate Professor and a Professor of Harbin Institute of Technology in 2012, 2016 and 2022, respectively. His research interests include nonlinear control and aircraft/spacecraft control. In these fields, he has led or participated as a core member in more than 10 national level scientific research projects, published over 50 journal papers, a monograph and a translation, and been authorized with 10 national invention patents. He is a reviewer for American Mathematical Review, a young editor of Aerospace Technology and Intelligence & Robotics, and a guest editor of Chinese Journal of Aeronautics and Journal of Systems Science and Complexity.


A singularity-free preassigned performance control approach


The preassigned performance control (PPC) methods have attracted considerable attention in recent years, however, most of the mainstream PPC methods utilize barrier functions and thus may suffer from the singularity problem of the control law under some unexpected conditions such as sensor faults. In this speech, a new robust PPC method without using barrier functions is introduced, which can completely avoid the singularity problem of the control law. The control design consists of three steps: firstly, a performance function is constructed to form the performance envelope according to the performance requirements; secondly, a preset trajectory (PT) is generated within the performance envelope; thirdly, a robust sliding mode control law is designed to make the actual tracking error evolve in a certain tubular area that contains the PT and is strictly within the performance envelope. Furthermore, the obtained result is extended to the case of adaptive PPC. The effectiveness and the advantages of the proposed PT based PPC method are illustrated by simulation results.


Professor Changzhong Pan

Hunan University of Science and Technology, China

Changzhong Pan received the B.S. degree in automation from Nanchang University, Nanchang, China, in 2006, and the M.Sc. and Ph.D. degrees in control science and engineering from Central South University, Changsha, China, in 2009 and 2013, respectively. From 2011 to 2013, he was a Visiting Scholar with Advanced Robotics and Intelligent Systems (ARIS) Laboratory, School of Engineering, University of Guelph, Guelph, Canada. He is currently a Professor and the Head of the Department of Automation at Hunan University of Science and Technology, Xiangtan, China. His research interests include flexible manipulators, intelligent robotics, underactuated mechanical systems, nonlinear and robust control with applications to mechatronic systems. He is a member of the Youth Editorial Board of Intelligence & Rbotics.


Advanced High-precision Intelligent Tracking Control of Flexible Manipulators with Uncertainties


Due to the distinctive features of light weight, good flexibility, high human-robot interaction safety, flexible manipulators (FMs) have a wide application prospect in the fields of industry, agriculture, medical service, aerospace, and so on. However, the flexible components are prone to produce elastic vibrations during the movements, especially in high-speed operations, which greatly affect the control accuracy of FMs. In addition, the model of FMs in practical applications may contain various nonlinear uncertainties including matched and mismatched disturbances. Advanced intelligent control has proved to be very efficient in dealing with complex uncertainties, where neural networks, adaptive control, and observers are often employed. This talk presents some new results of advanced control towards FMs with uncertainties, which covers vibration suppression, intelligent trajectory planning, active disturbance rejection control, and composite disturbance rejection control. On the one hand, based on artificial intelligence techniques, some new control theory and advanced control strategies with efficiency consideration are established for uncertain FMs. Simulations to typical flexible-joint and flexible-link manipulators are also included.


Assistant Professor Liguo Tan

Harbin Institute of Technology, China

1.September 2022-present: Head of Sino-Russian International Joint Laboratory on Intelligent Unmanned Systems, Harbin Institute of Technology.

2.December 2016-present: Assistant Researcher and Associate Researcher at the Laboratory for Space Environment and Physical Sciences (National Key Scientific Engineering), Harbin Institute of Technology.

3.May 2018-present: Master's and Ph.D. Supervisor in the Discipline of Control Science and Engineering, Harbin Institute of Technology.

4.May 2020: Awarded the Heilongjiang Province Excellent Youth Fund.

5.Principal Investigator for 12 national/provincial-level research projects.

6.Published more than 40 academic papers, including over 20 SCI papers.

7.Filed 53 national patents, 29 of which have been granted.

8.Authored 1 translated monograph.


A High-Precision Aero-Engine Bearing Fault Diagnosis Based on Spatial Enhancement Convolution and Vision Transformer


In aeronautical engineering, monitoring and diagnosing inter-shaft bearings of aero-engine is crucial for flight and life safety. Given that sensor signals are affected by strong noise environments, this study proposes a high-precision multi-sensor information fusion fault diagnosis method. The method is based on the lightweight spatial enhancement convolutional module with channel shuffling and combined with vision transformer (CSST-Net). Firstly, the original one-dimensional time-series signals acquired from multiple sensors are converted into two-dimensional time-frequency images using wavelet transform. The data is then fed into the model by utilizing data layer feature fusion. Subsequently, the interaction of information can be better facilitated by introducing channel shuffling operation into the convolution module. Additionally, the spatial enhancement algorithm leverages human visual perception properties to extract deeper fault features from the samples. Finally, the global information is extracted by the vision transformer to obtain the fault diagnosis results. By validating it on two different bearing datasets and comparing it with other classical and advanced fault diagnosis methods. The results show that the method proposed in this study is not only state-of-the-art but also extremely robust in strong noise environments. 


Assistant Professor Tangwen Yin

Shanghai Jiao Tong University, China

Dr. Tangwen Yin is an assistant professor at Shanghai Jiao Tong University. He is a member of the Intelligent Robot and Machine Vision Laboratory, Department of Automation. He received a B.S. in computer science from the National University of Defense Technology, an M.S. in precision instrument and machinery from the University of Shanghai for Science and Technology, and a Ph.D. in control science and engineering from Shanghai Jiao Tong University. He is an IEEE Senior Member, a CAA Senior Member, and a CAA Youth Academic Committee member. His research interest mainly lies in virtual physics artificial intelligence, algebraic geometry, harmonic measure, aviation and sonar electronics, flight control, and robotics multi-sensor signal processing. He has published eighteen academic papers, authorized five national invention patents, led one key project of the Aviation Science Foundation, two national defense and military industry projects, one joint research fund project of the Shanghai Commercial Aircraft System Engineering Science and Technology Innovation Center, and one project of the Ocean Equipment Forward Innovation Joint Fund.


Nonlinear Admissible Control Using Virtual Physics Artificial Intelligence


The forward designs of aircraft admissible control need integration of multidisciplinary domain knowledge, multi-source test data, and multi-echelon expert experience, and it plays an essential role in improving the safety and reliability of flight vehicles. This report will focus on the trustworthy math and data fusion mechanism and forward design models, aiming at challenges like knowledge embedding and data assimilation, full envelope and broad speed domain flight control, and performance bottleneck cause determination. Recent progress on nonlinear admissible control using virtual physics artificial intelligence includes a math and data fusion mechanism of admissible control, a nonlinear admissible control strategy, and an inverse attribution mechanism of multicoupling effect. In this way, it can not only break through the bottlenecks of knowledge embedding, data assimilation, admissible control, and inverse attribution technology but also overcome the availability limitations of cooperative decision-making in the forward design of aircraft admissible control from the perspective of robotics and automation. 


Assistant Professor Yaping Sun

Sichuan University, China

Yaping Sun received the Ph.D. degree in control science and engineering with the School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, China, in 2021. Since 2021, she has been a  assistant professor with the College of Electronic and Information Engineering, Sichuan University, Chengdu, China. Her research interests lie in the areas of switched multi-agent coordination control theory and its applications to autonomous robotics, unmanned ground vehicles, and unmanned aerial vehicles.


Consensus Tracking of Switched Heterogeneous Nonlinear Systems With Uncertain Target


This speech discusses a new dual-design framework that is used to investigate the consensus tracking problem of general switched heterogeneous nonlinear multi-agent systems (MASs) with an uncertain target. This framework includes a distributed observer and a distributed controller. A distributed observer that only uses the neighbors' information is constructed to estimate the target signal by designing a time-varying positive-define matrix function and mode-dependent observer gain matrices. A distributed controller is developed to solve the three consensus issues among the observer states, tracker states, and target states at the same time by utilizing the neighbors' observer states and designing additional mode-dependent feedback gain matrices. The distinct merit of the research is that the designed Lyapunov function is strict monotone decreasing at switching instants, discovering the positive effect of the switching on the consensus. As an application, heterogeneous switched Chua's circuits are provided to demonstrate this new framework.


Assistant Professor Timofey Kuzma

Moscow Polytechnic University, Russia

He was born on July 16, 1997, in the city of Orel. In 2016, he enrolled at Moscow Polytechnic University. Since 2017, he has been a member of the FDR Moscow team. He is responsible for the electrical and electronic systems of the team's vehicles, including selecting components, configuring control units, and manufacturing Mil-Spec wiring. From 2020 to 2021, he worked on the electrical system of  russian sports car Marussia GT. Additionally, since 2020 he has also worked as a racing engineer for the Sport GT team, working on both GT4 and GT3 vehicles.


Components and control logic of a hybrid sports prototype


— Determination of requirements for the components of a hybrid system.

— Analysis and selection of components based on world racing series standards.

— Development of control logic for the internal combustion engine and electric motor, depending on usage strategy.