CN108494514B

CN108494514B - Joint design method for predictive control and wireless short packet communication

Info

Publication number: CN108494514B
Application number: CN201810269710.2A
Authority: CN
Inventors: 赵国栋; 张馨月; 曹越; 童鑫; 陈智
Original assignee: University of Electronic Science and Technology of China
Current assignee: University of Electronic Science and Technology of China
Priority date: 2018-03-29
Filing date: 2018-03-29
Publication date: 2021-06-08
Anticipated expiration: 2038-03-29
Also published as: CN108494514A

Abstract

The invention belongs to the technical field of crossing of wireless communication and control, and particularly relates to a joint design method of predictive control and wireless short packet communication. The method can reduce communication energy consumption when carrying out joint optimization on the power, time-frequency resources and prediction length of the PPC system with a plurality of control loops. Through a plurality of control loops, the invention finds out the relation between the wireless communication consumption and the predicted length in the PPC system, thereby reducing the wireless communication cost.

Description

Joint design method for predictive control and wireless short packet communication

Technical Field

The invention belongs to the technical field of crossing of wireless communication and control, and particularly relates to a joint design method of predictive control and wireless short packet communication.

Background

The combination of traditional industrial control and wireless communication has mainly focused on two aspects: the control performance is optimized under a certain communication index. Another is to minimize the communication requirements under certain control parameters. It is assumed that the radio resource consumption monotonically increases with the increase of the communication demand, that is, the radio communication traffic brings more radio resource consumption, which is ubiquitous in the conventional network system. However, this assumption may not be valid in a cyber-physical system. This is mainly because the overall goal of control and wireless systems is to accomplish a control task rather than providing a bit stream. Therefore, in order to achieve a certain control performance, reducing the communication performance requirement may not result in a reduction in radio resource consumption.

The Packet Predictive Control (PPC) is a technique that reduces the reliability requirements by transmitting a packet containing future control information, often to cope with unreliable wireless links. In the PPC, the remote controller not only acquires the past state of the actuator and calculates the current control command, but also predicts the future state and the future control command. The remote controller packs the information and sends the information to the actuator, and if the wireless data packet is successfully received, the controller executes the current control command and caches the future control command; if a wireless data packet is lost or not arrived in time, the executor can autonomously execute the buffer command in the last data packet to continue the control task.

There are many studies to discuss PPC, but they ignore communication cost and control system in the wireless tight interaction process, i.e. the control and communication system is independently designed, so that the problem of energy consumption is common.

Disclosure of Invention

The present invention aims to solve the above problem, namely, how to jointly design the shared parameters of the control system and the wireless communication so as to minimize the wireless communication energy consumption while meeting the requirement of the control reliability of the control system, namely, the relationship between the predictive control system and the wireless communication cost in the PPC model.

For convenience in describing the contents of the present invention, terms and models used in the present invention will be described first:

packetized Predictive Control (PPC): an efficient technique for erasing a wireless channel for robust control.

Wireless Communication (Wireless Communication) is a Communication method for exchanging information using the property that an electromagnetic wave signal can propagate in free space.

Control System (CS, Control System): the management system with the self-target and the function consists of a control subject, a control object and a control medium.

Erasure Channels (EC, Erasure Channels): the channel is imperfect and sometimes "erased".

Gaussian Q function (GQ, Gaussian Q-function): right tail function of standard normal distribution.

Signal-to-Noise Ratio (SNR, Signal Noise Ratio): the ratio of signal power to noise power.

Time Delay (TD, Time Delay): there is a time difference in the received signals.

Additive White Gaussian Noise (AWGN), the most basic Noise and interference model, whose amplitude distribution follows a Gaussian distribution.

The technical scheme adopted by the invention is as follows:

as shown in fig. 1, which is a schematic diagram of a typical PPC system, a controller (controller) sends different commands to j actuators (activators) simultaneously through a wireless link. In the control loop, the controller includes a signal containing K for the actuator j_jA sequence of control commands, where k is 0 and k>0 indicates a control command for the current slot i and the future slot i + k, respectively. The wireless communication network then sends a packet u (i, j) to buffer j of current time slot i, where u (i, j) consists of a Kj command, which may be expressed as u (i, j) ═ u₀(i，j)、u₁(i，j)、...、u_K-1(i，j)]. Once buffer j successfully receives packet u (i, j), it will send the current command u₀(i, j) is handed over to executor j to execute, and the future command of K-1 is buffered. In the next time slot i + k, the buffer actually serves to prevent packet loss.

The principle of the invention is as follows:

by jointly designing the radio transmission power P₀(j) A wireless time-frequency resource n_jAnd controlling the prediction length K_jTo minimize wireless communication energy consumption and to guarantee a control interruption probability p_s. Minimizing total radio resource consumption of a predictive control PPC, wherein a constraint guarantees a control break probability p_sAnd finite time-frequency resources n_max。

The invention comprises the following steps:

if we define the control system interrupt as a controller with no executable commands, then K continues only at the controller_jThe failure to receive a wireless data packet occurs again, so the communication system needs to satisfy the following inequality.

p_e(j) Is that

Probability of packet loss, p_sIs the outage probability of the control system j. Short packet wireless communication is employed. Each packet has an H-bit header, and

is carrying K_jA control command. Therefore, the total number of bits of one packet is as follows.

N_j＝H+K_jL_j (2)

Let R_jFor short packet traffic capacity per unit bandwidth, the following expression can be obtained, namely

C＝log₂(1+γ_j) (4)

Variable gamma_jThe signal-to-noise ratio, C is the shannon capacity per unit bandwidth, and V is a channel dispersion coefficient. Here, Q^-1(. represents an inverse Gaussian Q function, n_jB.t stands for

The time-frequency resources of (2). Thus, there are:

optimizing the target:

optimizing transmission power P₀(j) Time-frequency resource n_jAnd a prediction length K_jTo minimize wireless communication energy consumption and to guarantee a control interruption probability p_s. Specifically, P is₀(j) Is shown as

) Then the signal-to-noise ratio (SNR) at buffer j is:

wherein N is₀Is the power spectral density of Additive White Gaussian Noise (AWGN). Then, the following radio resource consumption can be obtained:

this is the product of three basic resource elements of wireless communication, namely: power, frequency bandwidth, and time.

The following joint optimization problem is formulated:

C_j＝log₂(1+γ_j) (11)

the optimization problem is to use the total wireless resources of PPCMinimization of source consumption, wherein constraints guarantee a control interruption probability p_sAnd finite time-frequency resources n_max. To solve the optimization problem, an alternative optimization algorithm can be used to obtain the optimal solution and the comparison result, such as obtaining the optimal solution and the comparison result for a single controller, and the key points are to optimize the transmitting power P (j) and the prediction length K_j. According to the established model, the following can be solved mathematically:

will be optimal

The derivation of the objective function, instead of the regression equation, can yield:

at the same time, can be mathematically proved

Therefore, there are two cases of the original problem: if F'_k＝1>0, most preferably K ═ 1; if F'_k＝1<0, most preferably K is such that F'_kA solution of 0.

Subjecting the obtained product to

And

and replacing to obtain the minimum communication energy consumption expense E.

The method has the advantages that the method meets the requirement p of the reliability of the control system by jointly optimizing the prediction control length K, the wireless communication power distribution and the frequency spectrum resource distribution_sOn the premise of minimizing the wireless overhead E, a complete integrated system is obtained.

Drawings

FIG. 1 is a diagram of a system model of the present invention;

FIG. 2 is a diagram of simulation results of an embodiment.

Detailed Description

The technical solution of the present invention is described in detail below with reference to examples and drawings:

example (b):

in the first place, 4 control loops are set, i.e., j is 4, and an AWGN of the normalized power spectral density is considered. In the packet structure, the length of the header is 32 bits, and the length of each control command is 4, 8, 16, and 32 bits. Assuming the probability of interruption is the same for each platform, i.e. p_s＝10^-9。

The optimization model proposed by the invention is adopted to solve, as shown in fig. 2, the wireless energy cost and the maximum time-frequency resource n_maxThe relationship between them. The dotted line refers to the same time-frequency resource allocated to each platform and used as a baseline. The solid line is the result of the optimization according to the proposed method. As shown, the optimized result is much better than the platform averaging, indicating that the cooperative design of the control and wireless system effectively reduces the consumption of wireless resources. Secondly, it can be derived that the energy cost follows n_maxIs reduced because the spectral efficiency increases as the time-frequency resource n grows in short packet communications.

Therefore, when the method of the invention is used for carrying out the joint optimization on the power, the time-frequency resource and the prediction length of the PPC system with a plurality of control loops, the communication energy consumption can be reduced. Through a plurality of control loops, the invention finds out the relation between the wireless communication consumption and the predicted length in the PPC system. In order to reduce the cost of wireless communication, the invention jointly optimizes the transmission power, the time-frequency resource and the prediction length. The simulation result verifies the effectiveness of the method.

Claims

1. A joint design method of predictive control and wireless short packet communication is used for encapsulating a predictive control system to minimize the total wireless resource consumption of the system;

the encapsulation predictive controlThe system comprises a controller and J actuators, wherein the controller sends K to the J actuators through a wireless link_jA sequence of control commands, where k is 0 and k>0 indicating the control command of the current time slot i and the future time slot i + k respectively, and the wireless communication network transmits the data packet

To a buffer j of the current time slot i, where

From K_jA command component, denoted as

If buffer j successfully receives

Will be the current command u₀(i, j) is handed over to executor j to execute, and K is cached_j-1 future command;

the design method is characterized by comprising the following steps:

a. setting parameters:

the conditions for controlling the system without interruption are defined as follows:

wherein p is_e(j) Is that

Probability of packet loss, p_sIs the outage probability of the control system j;

short packet wireless communication is adopted, and the total number of bits of one packet is set as follows:

N_j＝H+K_jL_j

wherein H is the size of the packet header, L_jIndicates the required instruction size per slot of the executor j, and therefore

The payload of (A) is carrying K_jA control command;

let R_jShort packet traffic capacity per unit bandwidth:

C_j＝log₂(1+γ_j)

wherein the variable gamma_jTo the signal-to-noise ratio, C_jIs the Shannon capacity, V, per unit bandwidth_jIs a channel dispersion coefficient, Q^-1(. cndot.) represents an inverse Gaussian Q function;

by using n_j＝B_j·T_jIs represented by

Time-frequency resources of, B_jFor frequency bandwidth, T_jTime, therefore, there are:

b. establishing a model which minimizes wireless communication energy consumption and ensures that a control system is not interrupted:

by P₀(j) Is shown as

The snr at buffer j is then:

wherein N is₀A power spectral density that is additive white gaussian noise;

the wireless resource consumption model is established as follows:

the final model built is:

C_j＝log₂(1+γ_j)

n_maxis a finite time-frequency resource;

c. according to the number J of the actuators, realizing the time-frequency resource n by adopting an alternative optimization algorithm_maxThe allocation of (c).