Background
The inverted pendulum system has the characteristics of multivariable, strong coupling, nonlinearity, instability and the like, and becomes an ideal object for checking whether the control method has the capability of rapidly processing instability. In addition, stable control is carried out on the inverted pendulum system, and several key technical indexes in the control field can be effectively checked, wherein the key technical indexes comprise robustness, traceability, rapidity and the like. Therefore, designing different control methods for the inverted pendulum system to be stable has become a hot issue in the current control field.
Due to the fact that digital control technology is greatly improved and applied in the last decades, the inverted pendulum system achieves wireless network connection, and the control signal acquisition is completed based on sampling data of a system model. In the classical control framework based on sampled data, the control signal is updated according to a constant sampling period, which is called time sampling. While time sampling is advantageous for the integration and analysis of the controller, the time sampling approach is not ideal from a resource utilization perspective. In particular in network control systems, communication between sensors and control elements is achieved through limited network bandwidth. In this case, the communication between the sensors and the control components should be as small as possible to increase the utilization of network resources.
In recent years, an event-triggered communication mechanism has been proposed and proved to be a new strategy capable of effectively improving the utilization rate of network resources on the premise of ensuring the performance of a control system. Thus, the event-triggered strategy can be widely applied to the control problem of the inverted pendulum system. The invention patent "a design method of network control system controller based on event trigger" (CN202011118574.0) proposes a design method of network system controller based on event trigger mechanism, and the controller designed in the invention can ensure system stability under the condition of simultaneously considering introducing event trigger communication mechanism and packet loss. The invention patent 'electric power system control method based on event trigger dynamic trigger mechanism' (cn201710982346.x) designs an electric power system control method based on dynamic trigger mechanism, wherein the established event trigger mechanism is not only related to the current output and error, but also related to the output at the previous moment. The invention patent CN201910586233.7 discloses an output feedback controller of a neutral stable saturation system based on event triggering, and an established control mechanism can ensure the stability of the system when output saturation exists in the system. The local controller designed in the invention patent of networking control system and control method based on event trigger mechanism (CN201911245333.X) ensures that the networking control system has limited gain L2 stability and input feed-forward output feedback passivity.
It should be noted that in the above-mentioned patent of invention relating to the problem of networked system event-triggered control, a fixed gain controller is used throughout the event-triggered time, which makes the controller easy to implement. However, the control performance achieved using the fixed gain control method is not ideal. Therefore, how to design a more relaxed variable gain control method to ensure that the inverted pendulum system obtains better performance in a networked environment is a challenging problem, and related inventions do not exist at present.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
In this embodiment, the output feedback control of the inverted pendulum system of the trolley based on the event trigger mechanism includes the following steps:
step 1: for the inverted pendulum system, due to the instability of the inverted pendulum system, modeling is difficult, and after some factors are ignored, the linear primary inverted pendulum model can be regarded as a trolley inverted pendulum system as follows:
wherein x and theta are respectively the position coordinate of the trolley and the included angle between the pendulum and the vertical direction,
and
respectively the speed and the acceleration of the trolley,
and
angular velocity and angular acceleration, respectively, u being the control input, i.e. the control command, and y being the measurement output.
Modeling the trolley inverted pendulum system into a linear system, specifically as follows:
y(t)=Cx(t)
where x (t), y (t), u (t) are system state, measurement output and control input, respectively. A, B and C are system matrixes. In addition, (A, B) is controllable, and (A, C) is observable.
Step 2: establishing an event trigger communication mechanism, and designing a variable gain controller of the inverted pendulum system of the trolley;
(1) the event-triggered communication mechanism is established as follows:
wherein
t
k-1Which indicates the current moment of the data transmission,
which represents the state estimate that is transmitted,
representing the state estimate at the current time, e (t) representing the difference between the two.
The event threshold value is exponentially decreased, and the parameters respectively satisfy the following conditions: epsilon is more than 1, 0 is more than or equal to sigma and less1 and e
0Is more than or equal to 0. Furthermore, 0. ltoreq. t
0≤t
1≤…t
k≦ … is defined as the sequence of event-triggered times.
(2) Assuming the sensor has computational capability, the state estimate can be derived from a state observer through processing of the measured output, the observer being established as follows:
wherein
And
respectively an estimate of the state of the system and an estimate of the output, L being the observer gain matrix to be designed.
(3) Designing a controller with variable gain based on the established event trigger mechanism and the state observer, wherein the structure is as follows:
wherein K
kIs the controller gain to be designed, which is updated at the kth trigger time,
representing a positive integer. In addition, the data is sampled
Is transmitted to the controller side and the value is kept unchanged until the next time the sampled data arrives.
Defining new variables
And
an augmentation system of the following form is then obtained:
wherein
(4) The stability of the lyapunov function analysis system is established in the form:
V(t)=ξT(t)Pk(t)ξ(t)
wherein the Lyapunov function V (t) is discontinuous for t ≧ 0, a Lyapunov matrix Pk(t) is time-varying positive and satisfies the following inequality relationship:
wherein, λ (P)
k(t)) represents P
k-1The singular value of (a) is,
λ(P
k-1) And
respectively represent P
k-1Minimum and maximum singular values of, the symbols max, min and R
NRespectively, a maximum value, a minimum value, and a positive integer. In addition, P
k(t) is a continuous linear function, as follows:
wherein P isk-1> 0 and Pk+1> 0 denotes the time tk-1And tkIs given constant, δ ≧ 1 serves as a feasible solution for adjusting the linear matrix inequality.
In the embodiments of the present invention, for a given parameter e > 1, 0 ≦ σ < 1, e0Greater than or equal to 0 and delta greater than or equal to 1. If matrix P is presentk-1Observer gain matrix L and controller gain matrix KkAnd k is 1,2, …, satisfying the following inequality:
and
wherein, denotes the symmetric term of the matrix, I is the identity matrix, and phi is a positive constant.
The inverted pendulum system of the vehicle can achieve globally consistent and ultimately bounded stability, and all signals are contained within the following ranges:
(5) an event trigger mechanism gain, an observer gain, and a controller gain are determined.
1) Setting k to 1 and t to t0
By using the pole placement method, the following inequality is solved:
obtaining an initial parameter P0,K1And L. Further according to Pk(t) definition, assuming P0=P1。
2) Setting k 2 and t1
According to the parameter P obtained in step 1)1And L, by solvingThe following inequality:
to obtain K2。
3)t∈(t0,t1)
According to the parameters P obtained in 1) and 2)0,K1And L, solving the following inequality:
to obtain P2。
4) Setting k to 3 and t to t2
P obtained according to 3)2By solving the following inequality:
to obtain K3。
5)t∈(t1,t2)
According to the parameters P obtained in 1) and 2)1And K2By solving the following inequality:
to obtain P3。
Repeating the steps for N times.
And (6) ending.
According to 1) -N) in step 2, the event trigger mechanism gain, the observer gain and the controller gain matrix can be calculated.
And step 3: and (3) transmitting the control signal u established in the step (2) to an actuating mechanism of the inverted pendulum system of the trolley by the controller, thereby achieving the purpose of control.
The purpose of this embodiment is to ensure that the designed variable gain controller can make the inverted pendulum system of the trolley realize the global consistent bounded stability. In addition, the design method of the variable gain controller provided by the invention can obtain better system performance than the existing design method of the fixed gain controller.
The method is applied to an inverted pendulum system of a trolley to verify the effectiveness of the inverted pendulum system.
The inverted pendulum model of the vehicle is as follows:
wherein M is the trolley mass, M is the pendulum mass, l is the pendulum length, b is the friction of the trolley,
is the pendulum inertia. In addition, the selection of the respective parameters is as follows: m0.5 kg, b 0.1N/M/s, l 0.3M and I0.006 kg M
2。
Let e, σ 0.034 and e
00.01 is a parameter in the event trigger condition, and when the event trigger condition is established, the current state estimation value
Will be transmitted to the controller. Further, the definition ∈ ═ e, σ ═ 0.934, and ∈ e
00.0001 as a parameter in the non-trigger condition. Thus, it is possible to provideWith this method, a smaller number of triggers can be obtained. Without loss of generality, let t be
00, the first trigger occurs at
time t 00. By performing 1) in (5), the following initial gain matrix K can be obtained
1And L
K1=[17.0386 13.0877 -50.0520 -9.8150]
Then, based on the initial gain matrix K1And L, obtaining an event-triggered controller gain by performing 1) -N) in (5), as shown in fig. 2. The time and transmission interval of the data transmitted by the event trigger mechanism are shown in fig. 3. As can be seen from fig. 2-3, the gain of the event-triggered controller is varied and the event-triggered controller gain will switch when the event unit is triggered. In addition, fig. 4 plots the release data time and release interval for the event trigger mechanism obtained when using the fixed gain approach. As can be seen from fig. 3-4, the number of events triggered using the method of the present invention is 241, and the number of events triggered using the fixed gain method is 320. Fig. 5-8 show the system states obtained using the method of the present invention and the fixed gain method. Fig. 5-8 show that better control performance can be achieved with the method of the present invention with fewer triggers. Therefore, it can be seen from the simulation results that the design method disclosed in the present invention is effective.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.