1.College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, China 2.College of Civil Engineering and Architecture, Zhejiang Sci-Tech University, Hangzhou 310018, China
Fund Project:Project supported by the National Natural Science Foundation of China (Grant No. 61773348) and the Natural Science Foundation of Zhejiang Province, China (Grant No. LY16F030014)
Received Date:13 June 2019
Accepted Date:15 September 2019
Available Online:27 November 2019
Published Online:01 December 2019
Abstract:The problem of how to generate the Lorenz attractor from several nonlinear control systems is investigated in this paper. To be more precise, the conversions from the controlled Shimizu-Morioka system and the controlled Finance system to the Lorenz system are achieved by using the differential geometric control theory. For each case a scalar control input and a state transformation are proposed. The main approach of this paper is to convert all of those three-order systems into so called lower triangular forms which all have the same first two equations. Thus converting the controlled Shimizu-Morioka system or the controlled Finance system into the Lorenz attractor is feasible by choosing an appropriate scalar control input in the third equation of each of the two control systems. To this end, firstly, in order to use the tools of the differential geometry we construct a controlled Lorenz system by treating the vector field of the Lorenz attractor as the drift vector field and treating a linear vector field with three parameters as an input vector field. When those parameters are selected in a special manner, the conditions under which the controlled Lorenz system can be equivalently transformed into the lower triangular form are satisfied. Secondly, a state transformation, through which the controlled Lorenz system can be described as a lower triangular form, is obtained by a method like Gaussian elimination instead of solving three complicated partial differential equations. Employing several partial state transformations, choosing those three parameters and setting a scalar control input, we can reduce the equations of the controlled Lorenz system into its simplest lower triangular form. Thirdly, through two state transformations designed for the controlled Shimizu-Morioka system and the controlled Finance system respectively, the two control systems are converted into their lower triangular forms which are both similar to that of the Lorenz system in a way aforementioned. A smooth scalar controller is given to achieve the anti-control from the controlled Shimizu-Morioka system to the Lorenz attractor while another non-smooth scalar controller is designed to realize the generalized synchronization from the controlled Finance system to the Lorenz system no matter what the initial values of the two systems are. Finally, two numerical simulations demonstrate the control schemes designed in this paper. Keywords:chaos/ Lorenz attractor/ state transformation/ feedback
图 2 受控Shimizu-Morioka系统轨迹 Figure2. Trajectory of the controlled Shimizu-Morioka system
图 3 受控Shimizu-Morioka系统的标量控制输入 Figure3. Scale control input for the controlled Shimizu-Morioka system.
图 4 经状态变换${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$受控Shimizu-Morioka系统轨迹 Figure4. Trajectory of the controlled Shimizu-Morioka system via the state transformation ${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$.
图 6 受控Finance系统的标量控制输入 Figure6. Scale control input for the controlled Finance system
图 7 经状态变换${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$受控Finance系统轨迹 Figure7. Trajectory of the controlled Shimizu-Morioka system via the state transformation ${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$.
图 8 Lorenz系统轨迹与经状态变换${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$的受控Finance系统的误差 Figure8. Error between the trajectory of the Lorenz system and that of the controlled Finance system via the state transformation ${{{T}}^{ - 1}}({{\tau }}({{\zeta }}))$.