- We have developed a simple control algorithm for a class of linear disturbed systems with uncertainties. The main advantage is simplicity of the structure. The control law features one adjustable parameter and has low dynamic order. It guarantees robustness of closed system and exponentially converging of control error that is close to zero.
- The Laboratory has created and algorithm for adaptive control of objects with multiple inputs and multiple outputs in the presence of parametric uncertainty. The algorithm has been tested for a two-channel system. Conditions of applicability of the control law has been proposed that allows for exponential convergence of output rate to zero for the object. For a instable two-channel system computer modelling has been performed that confirmed efficiency of the proposed algorithm.
- An algorithm has been created for consensus control of dynamic network consisting of four nodes where every node is described by a non-linear differential equation with delay and unknown parameters. The algorithm provides synchronization of the network and compensation of uncontrolled disturbancies with set precision.
- Our researchers have synthesized an adaptive and robust control algorithm based on the method of sequential compensators for controlling a system with parametric and structural uncertainties (in cases when only the maximum possible relative order of the controlled object is known). To illustrate work of the algorithm, a program for controlling a mobile robot that uses computer vision has been created.
- A stabilizing algorithm has been developed that compensates delays for the class of nonlinear systems and provides asymptotic stability for a closed nonlinear systam in the presence of shifted harmonic disturbing signal. The created adaptive scheme allows to determine frequency and other parameters of outer disturbance that are used in the compensation circuit.
- A new approach has been discovered to stabilization of nonstable linear systems with big delays on inputs, unknown parameters and disturbance. We have proposed to use an algorithm based on predictor, the algorithm provides identification o parameters of the system and its stabilization. We have also analyzed a broader problem of evaluating and compensating an unknown disturbance. To demonstrate viability and efficiency of the proposed adaptive control we have conducted quantitative modeling
- Our researchers have proposed a method to control MIMO (multiple inputs, multiple outputs) systems on output. For example, we controlled a quadrocopter using sequential compensation by decomposition of the mathematical model into two parts: the first system is a static MIMO transform (the object is represented as a system of linear equations related to lifting power of engines and control inputs), while the second is represented as a number of SISO (single input, single output) channels.
- Our researchers have introduced a method to minimize data flow while monitoring a moving object through a communication channel. The method is based on usage of a binary adaptive coder. We have reviewed an example of monitoring the trajectory of a moving drone. A correlation has been found between load of the communication channel (volume of transmitted data over a time period) and precision of monitoring
- The laboratory has created a system to measure velocity of a vehicle moving along the road, the system is based on measurements from a sensor node placed on the roadside and consisting of an accelerometer and a magnetometer. Experiments at a test installation have shown high efficiency of the proposed method. However, the method should be enhanced for use on ordinary roads.
- We have solved the problem of enhancing frequency identification in single frequency harmonic signals. The proposed cascade consists of adaptive bandpass filter tuned to evaluate frequency provided by the proposed identification algorithm. Stability of the cascade has been studied and limitatons of the trajectory has been proven by analysis of the Lyapunov function and finding allowances of the identification algorithm.
- Our researchers have created phase regulator for a population of oscillating systems. The proposed approach is based on the model of phase response curve for an isolated oscillator (regulated first order model for a system with infinitely small input linearized along the limit cycle of the system). Our team has proven that phase regulation is also achieved for the initial nonlinear system
- We have proposed a new robust, nonlinear globally convergen position observer for a fixed synchronous engine with permanent magnets. THe key feature of the porposed regulator is that it only requires knowledge of stator resistance and inductivity while mechanical parameters and magnetic flow may remain unknown. Due to new re-parametrization of engine dynamics, only two parameters of the engine are evaluated by the regressor including filtered voltage and current during normal operation of the engine
- The Laboratory has proposed a new class of devices for evaluation of synchronous engines with permanent magnets. Thanks to usage of a new representation of the engine and suitable filters we have obtained a new solution for two important problems: evaluation of stator resistance and inductivity and evaluation of flow with unknown electric parameters. None of the schemes require knowledge of mechanical parameters and magnetic flow.
- Our researchers have developed a method for finite control of a chain of integrators. The proposed approach provides finite stability and robustness against internal disturbance
- We have studied conditions of ISS stability for systems with several positions of equillibrium abd time delays based on on the Razumikhin–Lyapunov functional. Robustness and stability of the system against delays in tje feedbak channel has been shown.
- Our researchers have developed a control algorithm for second order nonlinear systems in the presence of parametric uncertainty and external limited disturbances. The algorithm is built upon the backstepping method and supplemental circuit. This provided stability to the system compared to the classical method of backstepping.
- The problem of stabilizing electric generators has been solved. The proposed algorithm implements the system of monitoring the reference value of phase by measuring relative angles of rotation of the generator's rotors and regulating currents at their windings
- We have developd a modified finite control algorithm based on usage of the Lyapunov function and properties of homogenous systems
- Our Laboratory has analyzed the algorithm of solving the problem signal uncertainty in constructing a sequential compensator in the problem of controlling piezoelectric drive to check its working capacity and efficiency
- The researchers have developed a method to synthesize algorithms of sensorless control of non-salient pole synchronous engines with permanent magnets using the proposed position observers. The control system in combination with a nonlinear obeserver has been compared with a modern industrial regulator. We have shown that the proposed sensorless regulator wuth nonlinear observer allows to significantly improve characteristics of the closed circuit while the engine is working at bpth low and high speed.
- Our staff has approbated an adaptive system for monitoring multi-harmonic signals in the presence of input delay causd by using the Internet as a means of communication
- We have developed algorithms of robust control for nonlinear systems that provide convergence of all the trajectories of a closed system into a compact set in the presence of delay in the control channel. In the absence of delay, finite stability of the closed system is guaranteed. The achieved results are also holds for variable delays and several delays.
- Our researchers have reviewed the problem of regulating outputs of multichannel systems with harmonic input impacts and harmonic uncertainties. We have reviewed the problem of stability with using properties of «robust» minimal phase. Our team has developed a regulator based on the method of classical method of internal model together with a circuit of adaptive tuning. The regulator provides asymptotic convergence of output variables to positions of equilibrium. Experimental approbation of the proposed algorithm to control a surface vessel
- The Laboratory has developed an algorithm allowing to filter payload from chaotic signal to evaluate parameters of the system. We have also created an observer that uses only output signal of a chaotic system in the circumstances of complete parametric uncertainty of the model
- An algorithm for planning routes of industrial robots has been developed. It is based on approximation of the set circuit using curves. The solution allows to reduce the number of support points, to reduce code size and computational costs, to increase quality of operations as well as to simplify programming complex motions. Experiments on six-section manipulator with rotating joints have been conducted.
- An algorithm for evaluation of unknown parameters of Solar element using iterational the Newton-Raphson method has been created
- We have proposed an method to identify the value of maximum power of a photoelectric source in various atmospheric circumstances based on P&O and INC methods.
- A method has been proposed for identifying properties of solar batteries using the prcedure of dynamic expansion of regressor
Implemented results of research:
We have found practical applications for: the algorithm of identification of parameters of multiharmonic signal (for filtering naviation data), method of controlling objects with uncertain parameters in disturbances (for distributing supports and dynamic positioning of a vessel at a point)
Education and career development:
- We have developed and launched 4 masters programs: «Digital control in modern equipment» (2014), «Industrial robotics» (2016), «Adaptive and robust control of nonlinear systems» (2017), «Sensorless control» (2018) as well a pedagogical staff training program «Modern technologies of control systems synthesis» (2015).
- Our staff have developed 3 doctoral dissertations, 12 candidate dissertations, 25 masters dissertations and 8 bachelor dissertations
- We have published the textbook «Designing intelligent control systems for house automation. Elements of theory and practice».
- 10 professionals have completed the «Adaptive and nonlinear control systems» training course at the Laboratory. The professionals, young scientists and tutors represented the University of Leeds (UK), Trapeznikov Institute of Control Sciences of the Russian Academy of Sciences (Russia), Astrakhan State Technical University (Russia), Gubkin Russian State University of Oil and Gas (Russia),Peter the Great St. Petersburg Polytechnic University
(Russia)
Other results:
- We have developed specialized software for computers and robotic complexes: a program to measure amplitude frequency response sensibility functions, program fir controlling mobile robots using gestures, voice and mimics, program for graphical analysis of transition processes in linear dynamic systems of first and second order, program for stabilization of the Darwin-OP robot, program for controlling motion of the Darwin-OP robot using computer vision, program to control the Pololu Dual VNH5019 driver, Surfer software to evaluate position and to stabilize humanoid robots on an inclined surface and callculate coordinates of centers of inertia of its parts, program for measuring angular position of humanoid robots, program for decentralized control of rotary-wing aircraft
- We have calculated a test ridge to probe suspension of automotive vehicles
- Our team has developed robust stabilization algorithmms for a two-legged robot in the standing position Snowboarder and Snowboarder 2.1
- Sequential compensator in the presence of quantization
- We have developed a mobile robot with horizontal and vertical movement as well as a robotic grasping hand
- We have proposed, developed software and tested an algorithm for planning trajectories of a humanoid robot balancing in the standing position
- We have proposed, developed software and tested an algorithm for planning trajectories of a humanoid robot based on computer vision data
Collaborations:
- Institute of Machine Sciences of the Russian Academy of Sciences (Russia), Centre National de la Recherche Scientifique (France), University Valenciennes (France), Institute national de recherche en informatique et en automatiq (France): joint scientific events, research and student exchanges
