A joint design method for control transmission based on full-loop information age

Abstract

The present invention discloses a control transmission joint design method based on the full loop information age, and relates to the technical field of the full loop information age. The present invention performs control and transmission joint design based on the full loop information age, and the control strategy of the controller and the trigger strategy of the event trigger are both based on the full loop information age. The time sensitive network is a deterministic transmission mechanism, and the routing and scheduling schemes are obtained by modeling and optimizing based on the full loop information age, so as to ensure the deterministic and reliable transmission of the data stream transmitted in the network and improve the overall performance of the industrial information physical system.

Description

Technical Field

[0001] The present invention relates to the technical field of full-loop information age, and in particular to a control transmission joint design method based on full-loop information age. Background Art

[0002] Industrial cyber-physical system is a multi-dimensional complex system that integrates computing, network and physical environment. It integrates advanced computing, transmission and control technologies, focuses on the close combination and coordination of computing resources and physical resources, improves the operating efficiency of industrial automation systems, smart grids, smart logistics and digital twins, and is one of the trends of future industrial development. Since different subsystems in industrial cyber-physical systems are tightly coupled and influence each other in various ways through network transmission, it is difficult to improve the overall performance of industrial cyber-physical systems by optimizing the control and transmission systems separately. For example, reducing the number of data transmissions in industrial cyber-physical systems can reduce the network burden, but the resulting information loss may seriously reduce the control performance. Therefore, the joint design of control and transmission strategies is crucial to improve the overall performance of industrial cyber-physical systems.

[0003] Since latency is a key indicator of the state of the transmission system in industrial cyber-physical systems, and it can also affect the stability of the control system in industrial cyber-physical systems, most of the existing control transmission joint design strategies are studied based on latency. On the other hand, event triggering is widely used in the design of industrial cyber-physical systems because it can better optimize computing and transmission resources than time triggering. However, latency only reflects the time that data takes in the network link or controller calculation process, and cannot be used to measure the delay caused by event triggering to information. Therefore, for industrial cyber-physical systems that include control systems and transmission systems, considering latency alone cannot fully measure the freshness of information in industrial cyber-physical systems. Designing the control transmission joint design problem with event triggering based on latency will affect the improvement of the overall system performance.

[0004] As an emerging indicator, the full-loop information age describes the freshness of information in the entire industrial cyber-physical system loop, which takes into account the time that information has experienced in the trigger-controller transmission link (uplink), the controller-actuator transmission link (downlink) and the event trigger. Through the full-loop information age, the control system and transmission system in the industrial cyber-physical system can obtain each other's current status and requirements in a timely manner, which facilitates the remote coordination and resource allocation of the control system and the transmission system. Therefore, under the premise of ensuring the accuracy of calculation, the control transmission joint design based on the full-loop information age can reduce the transmission burden of the network system as much as possible while ensuring the stability of the control system, thereby improving the overall performance of the industrial cyber-physical system.

[0005] In the research on the design of industrial cyber-physical systems, few existing related patents can jointly design the control system and the transmission system. The patent application number is CN202210172922.5, and its name is a control method for a cyber-physical system controller based on an event-triggered mechanism. It combines the characteristics of network-based data transmission of cyber-physical systems and designs a control strategy based on a new event-triggered mechanism under deception attacks. The system can achieve rapid compensation when an actuator fails, improve the system control performance, and the system has a higher data packet sending rate when it is attacked and there are faults. When the system tends to be stable, the data packet sending rate decreases, reducing the communication burden and saving bandwidth resources. However, the focus of the patent is the design of the controller, through which network resources are optimized, and there is no design for the transmission system, which limits the further improvement of the performance of industrial cyber-physical systems. The patent application number is CN201521110828.9, and its name is a remote data transmission system for debugging electric vehicle motor controllers. The system includes a power module, a dynamometer motor and controller, a motor under test and a controller. The power module provides the working power of the motor controller through the power switch or ignition switch of the motor controller. It also includes a communication gateway, a relay and an intelligent terminal. The communication gateway includes devices such as a CAN interface and a network card. The data transmission system overcomes the defects of traditional motor controller debugging and realizes the remote transmission of data during the debugging of the motor controller. It is convenient to wire, avoids the long-term duty of the tester, and improves the debugging and R&D efficiency. However, the focus of the patent is on the design of the transmission system. It improves the debugging efficiency of the controller by developing transmission-related hardware systems, and does not design for control, which limits the further improvement of the debugging and R&D efficiency. The patent application number is CN202110434832.4, and the name is a controller and data transmission design method for DSP and ZYNQ architecture. The data transmission method between the ZYNQ control board and the DSP control board provided by the patent realizes real-time and effective transmission of data between the DSP and the ZYNQ by configuring the data transmission line and the address transmission line, which is convenient for configuration and expansion according to actual control needs. However, the control and transmission joint design object provided by the patent is only at the control board level, and cannot be extended to the design of large-scale, multi-flow, and multi-subsystem industrial cyber-physical systems. In summary, there are currently few control and transmission joint design solutions related to cyber-physical systems, and a complete set of design solutions is needed to improve the overall performance of cyber-physical systems.

[0006] On the other hand, in the research and application of information age, on the one hand, the existing information age does not fully describe the freshness of information, and on the other hand, there is little joint design work on control transmission based on information age. The patent application number is CN202110119244.1, and its name is a wireless network communication resource scheduling method and system based on information age. The wireless network communication resource scheduling method and system based on information age provided by the patent provide a basis for the transmission fairness between links, while improving the real-time performance and information timeliness of the overall system. The patent application number is CN202011077538.4, and its name is a multi-channel wireless network scheduling method supporting information age optimization. This patent schedules multiple channels, improves network resource utilization, and optimizes the average information age of the network; at the same time, since the target node does not need to obtain the data packet generation time of the source node in advance, it also reduces the energy consumption caused by the interaction between nodes. However, the above patent's definition of information age is relatively simple, which only expresses the difference between the time when the target node generates the latest data packet from the source node and the current time, without defining the information age based on the process that the data goes through from generation to final use. As a result, the information age cannot be designed from the underlying mechanism, which limits the use of the information age.

[0007] In summary, it is necessary to redefine the information age based on the process that data goes through in the cyber-physical system loop, and to jointly design the control and transmission systems in the cyber-physical system based on the information age to improve the overall performance.

[0008] The existing technology has a relatively simple definition of information age. Simply subtracting the current time from the time when the data was generated cannot fully reflect the data transmission process, nor can it design the information age. It is necessary to give the information age more mechanisms to describe the freshness of the information. The existing technology optimizes the utilization of network system resources based on information age, but does not design a corresponding transmission mechanism to ensure the accuracy of information age calculation. Calculating information age under mathematical expectation cannot accurately guide the optimization design of the system. The existing technology is mostly separate for the design of control and transmission in industrial information-physical systems, which only improves the control system or the transmission system, but does not analyze the mutual coupling relationship between control and transmission and jointly design them.

[0009] Therefore, technicians in this field are committed to developing a control transmission joint design method based on the full-loop information age to ensure the rapidity and stability of the control system, and to reduce the amount of data transmission in the network as much as possible, reduce the network burden, and improve the overall performance of the industrial network system. Summary of the invention

[0010] In view of the above defects of the prior art, the technical problem to be solved by the present invention is to propose a new definition of information age, enrich the description content of information age for information freshness, and enable it to be used as a key joint index for the joint design of industrial cyber-physical system control and transmission. By adopting a suitable network transmission mechanism, a corresponding transmission model is established and a transmission algorithm is designed to ensure the accuracy of information age calculation. The information age is used to characterize the mutual influence relationship between the control and transmission systems, and the joint design of the control and transmission systems is carried out based on the information age to improve the overall performance of the industrial cyber-physical system.

[0011] To achieve the above object, the present invention provides a joint control and transmission design method based on full-loop information age, including the following steps:

[0012] Step 1: For the industrial cyber-physical system, for a given network topology and the control period of the data flow of each subsystem, generate a feasible deterministic routing and scheduling scheme S; assume there are M such schemes in total; initialize the scheduling scheme counting variable n1; the industrial cyber-physical system contains N subsystems in total, initialize the subsystem counting variable n2; the design duration of the industrial cyber-physical system is T, initialize the design start time k of subsystem n2.

[0013] Step 2: If k < T - τ, jump to Step 3;

[0014] If K ≥ T - τ and n2 < N, then let temp = n2 + 1, n2 = temp,, k = 0, and return to Step 2;

[0015] If k ≥ T - τ and n2 ≥ N and n1 < M, then determine whether the full-loop information age corresponding to the n1 scheduling scheme satisfies the control stability condition. If so, calculate the corresponding transmission cost, control cost, and total cost under the n1 scheduling scheme:

[0016]

[0017] And let temp = n1 + 1, n1 = temp, n2 = 1, k = 0, and return to Step 2; otherwise, let temp = n1 + 1, n1 = temp, n2 = 1, k = 0, and return to Step 2;

[0018] If k ≥ T - τ and n2 ≥ N and n1 ≥ M, jump to Step 5;

[0019] Step 3: For the n1th subsystem, for the n2th delay set, calculate the corresponding at time k Judge Whether it is less than 0. If so, then If not, then let

[0020] Step 4: According to and The value result is based on the full loop information age Δ k+τ Calculate the optimal control solution Let temp = k + 1, K = temp, and return to step 2;

[0021] Step 5: For n1∈[1,M], compare the corresponding Select the smallest As the optimization result, it is recorded as the total cost of optimal control transmission Ω * ; The corresponding routing scheduling scheme S, control strategy c, and event triggering scheme γ are taken as the control transmission joint design results, denoted as S * 、c * , γ * and

[0022] Furthermore, in step 1, the industrial cyber-physical system includes a controlled object, a sensor, an actuator, an event trigger based on the full-loop information age, a controller based on the full-loop information age, and a time-sensitive network; the state information x of the controlled object k The sensor collects and transmits it to the event trigger, which determines the current x based on the age of the full loop information. k Whether to trigger; if triggered, x k Transmitted to the uplink time-sensitive network; if not triggered, wait for receiving the next moment x k+1 , and judge whether it is triggered or not; x k After entering the uplink time-sensitive network, it is transmitted to the controller according to the routing and scheduling strategy designed offline. The corresponding delay of the process is τ tc ; The controller calculates the control input u based on the full loop information age k , and transmitted to the downlink time-sensitive network; u k In the downlink time-sensitive network, the routing and scheduling strategy designed offline is transmitted to the executor, and the corresponding delay τ ca ; Total delay τ = τ tc +τ ca ; The operation of the final actuator causes the controlled object to update its state and start the update process at the next moment.

[0023] Furthermore, in step 1, the time-sensitive network is designed based on the IEEE 802.1Qbv standard.

[0024] Furthermore, in step 1, in an offline state, a gated queue for switch transmission is formed by designing a deterministic transmission constraint set; in actual operation, each transmission port of the switch is periodically opened and closed according to the time specified by the gated queue designed offline to ensure the determinism of data transmission.

[0025] Furthermore, the transmission constraint set includes topology constraints, contention-free constraints, transposition constraints, loop-free constraints and routing constraints.

[0026] Furthermore, in step 2, the control stability condition is a constraint on the average full-loop information age. When the average full-loop information age is less than a stability threshold, the system is stable.

[0027] Further, in step 3, the event trigger based on the full loop information age is based on Δ + The value of is used to trigger the operation, specifically:

[0028]

[0029] Where ξ is the cost of opening and closing the gate of the time-sensitive switch once, It is the total number of times that all switch gates are opened and closed during the transmission of data in a time-sensitive network. in is the system matrix, ω is the environmental noise; Θ k+τ =L k+τ Λ k+τ L k+τ is the solution-dependent variable expression of the Riccati equation; when Δ + When the value of is greater than 0, the trigger occurs; when Δ + When the value is less than 0, the trigger does not occur.

[0030] Furthermore, in step 3, the definition of the full loop information age is

[0031] Δ k+τ =(1-δ k )(Δ k+τ-1 +1)+δ k τ

[0032] Among them, δ k is the event trigger variable, τ is the total network transmission delay; if the event trigger occurs at time k, the value of the full loop information age is the total loop delay, that is, Δ k+τ =τ; if the event trigger does not occur at time k, the value of the full loop information age increases by 1 compared to its value at the previous time, that is, Δ k+τ =Δ k+τ-1 +1; δ k The value of Δ +Decide when Δ + When the value of is greater than 0, the trigger occurs, δ k =1; when Δ + When the value is less than 0, the trigger does not occur. k =0.

[0033] Furthermore, in step 4, the control strategy based on the full-loop information age utilizes the LQR algorithm and defines new state information and input, thereby obtaining the optimal control decision.

[0034] Furthermore, in step 5, the system compares the total costs corresponding to the control transmission joint design based on the full-loop information age under each routing scheduling strategy, and takes the solution with the smallest total cost as the optimal solution in the feasible solution set, including the optimal routing scheduling strategy, the optimal control strategy and the optimal event triggering strategy.

[0035] In a preferred embodiment of the present invention, the purpose of the present invention is to define a new concept of information age - full-loop information age, and perform control and transmission joint design based on the full-loop information age, so as to utilize the joint design method and system to improve the overall performance of industrial information-physical systems.

[0036] To achieve the above object, the present invention adopts the following technical solution:

[0037] The system as a whole is composed of a controlled object, a sensor, an actuator, a full-loop information age event trigger, a full-loop information age controller, and a time-sensitive network, wherein the time-sensitive network includes several switches. Among them, the control strategy of the controller and the trigger strategy of the event trigger are based on the full-loop information age. The time-sensitive network is a deterministic transmission mechanism. By modeling and optimizing the solution based on the full-loop information age to obtain the routing and scheduling scheme, it can ensure the deterministic and reliable transmission of the data stream transmitted in the network.

[0038] A control and transmission joint design method based on full-loop information age includes the following steps:

[0039] Step 1: The controlled object updates the state information x at time k k After that, the sensor will send the measured state information x k Passed to the event trigger.

[0040] Step 2: Event trigger receives x k Afterwards, it is determined whether to trigger the data according to the trigger strategy based on the age of the full-loop information.

[0041] Step 3: If the event trigger is not triggered, jump to step 6; if triggered, the state information x kEntering the time-sensitive network, according to the routing and scheduling algorithm based on the full-loop information age under the time-sensitive network, after τ tc After a certain time, it is transmitted to the controller.

[0042] Step 4: The controller receives x k Then, the control input u is calculated based on the control strategy based on the full loop information age. k .

[0043] Step 5: Control input u k Entering the time-sensitive network, according to the routing and scheduling algorithm based on the full-loop information age under the time-sensitive network, after τ ca After the time is reached, it is transmitted to the actuator, and the actuator executes according to u k Operate the controlled object. Total network transmission delay τ = τ tc +τ ca .

[0044] Step 6: The controlled object updates the state to obtain x k+1 , let temp = k + 1, k = temp, and return to the first step.

[0045] In particular, in step 1, the system includes N subsystems, each subsystem includes 1 controlled object and 1 event trigger, the controlled object and the event trigger correspond one to one, and the state information of different subsystems does not affect each other during the event triggering stage.

[0046] In particular, in step 2, the triggering operation of the event trigger is reflected in the value of the event trigger variable, and the event trigger variable is defined as follows:

[0047]

[0048] When an event trigger occurs, δ k The value is 1; when the event trigger does not occur, δ k The value is 0.

[0049] In particular, in step 2, the full loop information age is defined as:

[0050] Δ k+τ =(1-δ k )(Δ k+τ+1 +1)+δ k τ

[0051] Among them, δ k is the event trigger variable, and τ is the total network transmission delay. If the event trigger occurs at time k, the value of the full loop information age is the total loop delay, that is, Δ k+τ =τ; if the event trigger does not occur at time k, the value of the full loop information age increases by 1 compared to its value at the previous time, that is, Δk+τ =Δ k+τ-1 +1.

[0052] In particular, in step 2, the optimal event trigger variable group is constructed based on the full loop information age, specifically:

[0053]

[0054] Where ξ is the cost of opening and closing the gate of the time-sensitive switch once, It is the total number of times that all switch gates are opened and closed during the transmission of data in a time-sensitive network. in is the system matrix, ω is the environmental noise; Θ k+τ =L k+τ Λ k+τ L k+τ is the solution-dependent variable expression of the Riccati equation; when Δ + When the value of is greater than 0, the trigger occurs; when Δ + When the value is less than 0, the trigger does not occur.

[0055] In particular, in step 3, the time-sensitive network is a deterministic network. Given the network topology and data flow cycle and other characteristics, the optimal routing and scheduling solution is obtained by offline design of scheduling optimization objective function and related constraint sets to ensure deterministic transmission. After the optimal routing and scheduling solution is given, the transmission time of data flow in the time-sensitive network is τ tc , τ ca The data transmitted in step 3 is the state information x. After the time-sensitive network scheduling scheme is optimized offline, the time required for the state information transmission is τ. tc Known.

[0056] In particular, in step 3, the design process of the time-sensitive network routing and scheduling algorithm based on the full-loop information age includes the construction of the optimization problem and the constraint set, and the solution of the optimization problem. The objective function of the optimization problem is as follows:

[0057]

[0058] in,

[0059]

[0060] is the total cost of control and transmission, λ1 and λ2 are weight coefficients, J is the control cost function, is the transmission cost function, c is the control strategy, and γ is the transmission strategy. The optimization goal of the time-sensitive network routing and scheduling algorithm based on the full-loop information age is to minimize Ω. Among them, J is defined as

[0061]

[0062] Among them, Q and R are weight coefficient matrices. This definition is the control cost form of LQR optimal control.

[0063] The definition is

[0064]

[0065] Where ξ is the cost of opening and closing the gate of the time-sensitive switch once, is the total number of times that all switch gates are opened and closed during the transmission of data in the time-sensitive network. con The definition means the total cost of switching on and off of a time-sensitive network switch. The first term is the theoretical number of periodic switching times of the switch within time T, and the second term is the amount of savings in the switching cost caused by the absence of an event trigger.

[0066] In particular, in step 3, the design process of the time-sensitive network routing and scheduling algorithm based on the full-loop information age includes the construction of the optimization problem and the constraint set, and the solution of the optimization problem. The constraint set includes the relevant network transmission constraints defined by the IEEE 802.1Qbv standard, the event triggering constraints mentioned in step 2, and the control algorithm constraints mentioned in step 4. In an offline case, the optimization problem is solved based on the preprocessing and step-by-step solution algorithm.

[0067] In particular, in step 4, the control algorithm based on the full-loop information age adopts the LQR algorithm, which gives the elimination of the input delay τ ca The optimal input calculation scheme of the influence is given, and the stability condition of the control system is given in combination with the average full-loop information age. ca The existence of , the stability of the original system is affected. The original system equation After discretization,

[0068] x k+1 =Ax k +B1u k-α+1 +B2u k-α ω k

[0069] Among them, τ ca =αp con -β, α is an integer, β∈[0,p con ]. A, B1, B2 are system matrices, ω k is a noise function. At this time, the information age of the whole loop Δ=τ caIn order to eliminate the impact of input delay, a new state quantity is defined based on the full loop information age

[0070]

[0071] Then the original system equation is transformed into a time-delay-free form in

[0072] In particular, in step 4, the control algorithm based on the full-loop information age adopts the LQR algorithm, which gives the elimination of the input delay τ ca The optimal input calculation scheme for the influence of is given, and the stability condition of the control system is given in combination with the average full-loop information age. The optimal control input is

[0073]

[0074] in, where Y k+1 is the solution of the following Riccati equation

[0075]

[0076]

[0077] in,

[0078] In particular, in step 4, the control algorithm based on the full-loop information age adopts the LQR algorithm, which gives the elimination of the input delay τ ca The optimal input calculation scheme for the influence of is given, and the control system stability condition is given in combination with the average full-loop information age. The system stability condition is

[0079]

[0080] in, is the average full-loop information age, ζ is the maximum value of the matrix trace, ∈ is the system dimension, σ is the noise variance, x0 is the initial state, is the weight coefficient matrix. When the average full-loop information age is less than the above expression, the system is stable.

[0081] In particular, in step 5, the process of transmitting the control input to the actuator via the time-sensitive network is independent and does not conflict with the process of transmitting the state information to the controller via the time-sensitive network.

[0082] Compared with the prior art, the present invention has the following obvious substantial features and significant advantages:

[0083] 1. A new definition of information age is proposed. The full-loop information age under this definition takes into account the entire process of data transmission in industrial cyber-physical systems, including event triggering, network transmission from event trigger to controller, and network transmission from controller to actuator. It also quantitatively characterizes the delay in the corresponding process, so that it can fully characterize the freshness of information.

[0084] 2. A deterministic network transmission mechanism is adopted, and a corresponding transmission model is established and a transmission algorithm based on the full-loop information age is designed to ensure the reliability and determinism of data transmission in industrial cyber-physical systems. Compared with the existing technology, this method enables the information age to be accurately calculated, ensuring the feasibility of the control transmission joint design based on the information age.

[0085] 3. The full-loop information age is used to characterize the mutual influence between the control and transmission systems. The value of the full-loop information age is affected by the scheduling scheme of deterministic transmission, the time when the event triggering based on the full-loop information age occurs, and the control decision based on the full-loop information age is also affected. In addition, this technology provides the control stability conditions based on the full-loop information age. By constructing the optimization goal of the overall cost of control transmission and taking the routing scheduling, event triggering, and control strategy based on the full-loop information age as a joint constraint set, the overall performance of the industrial information-physical system can be further improved.

[0086] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 It is a preferred embodiment of the present invention, based on the control transmission joint design architecture of the full loop information age;

[0088] Figure 2 is a physical topology connection diagram of a preferred embodiment of the present invention;

[0089] Figure 3 It is a full loop information age growth curve of a preferred embodiment of the present invention;

[0090] Figure 4 It is a control transmission joint design algorithm flow based on the full-loop information age of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0091] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.

[0092] In the drawings, components with the same structure are denoted by the same numerical labels, and components with similar structures or functions everywhere are denoted by similar numerical labels. The dimensions and thicknesses of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the dimensions and thicknesses of each component. To make the illustration clearer, the thicknesses of some components in the drawings are appropriately exaggerated in some places.

[0093] A control transmission joint design method and system based on the age of information of the full loop of the present invention includes the following steps:

[0094] The first step: For an industrial cyber-physical system, for a given network topology and the control period of the data flow of each subsystem, generate a set of feasible deterministic routing and scheduling schemes S. Assume that there are M such schemes in total. Initialize the scheduling scheme counting variable n1 = 1. The industrial cyber-physical system contains N subsystems in total, and initialize the subsystem counting variable n2 = 1. The design duration of the industrial cyber-physical system is T, and initialize the design start time k = 0 of subsystem k2.

[0095] The second step: If k < T - τ, jump to the third step.

[0096] If k ≥ T - τ and n2 < N, then let temp = n2 + 1, n2 = temp, k = 0, and return to the second step.

[0097] If k ≥ T - τ and n2 ≥ N and n1 < M, then judge whether the age of information of the full loop corresponding to the n1 scheduling scheme meets the control stability condition. If so, calculate the corresponding transmission cost, control cost and total cost under the n1 scheduling scheme:

[0098]

[0099] And let temp = n1 + 1, n1 = temp, n2 = 1, k = 0, and return to the second step. Otherwise, let temp = n1 + 1, n1 = temp, n2 = 1, k = 0, and return to the second step.

[0100] If k ≥ T - τ and n2 ≥ N and n1 ≥ M, then jump to the fifth step.

[0101] The third step: For the n1th subsystem, for the n2th delay set, calculate the corresponding at time k Judge Whether it is less than 0. If so, then If not, then let

[0102] The fourth step: According to And The value result of, based on the age of information of the full loop Δk+τ Calculate the optimal control solution Let temp = k + 1, k = temp, and return to the second step.

[0103] Step 5: For n1∈[1,M], compare the corresponding Select the smallest As the optimization result, it is recorded as the total cost of optimal control transmission Ω * The corresponding routing scheduling scheme S, control strategy c, and event triggering scheme γ are taken as the control transmission joint design results, denoted as S * 、c * , γ * and

[0104]

[0105] In the first step, the industrial cyber-physical system includes a controlled object, a sensor, an actuator, an event trigger based on the full-loop information age, a controller based on the full-loop information age, and a time-sensitive network. k The sensor collects and transmits it to the event trigger, which determines the current x based on the age of the full loop information. k Whether to trigger. If triggered, x k Transmitted to the uplink time-sensitive network; if not triggered, wait for receiving the next moment x k+1 , and judge whether it is triggered or not. k After entering the uplink time-sensitive network, it is transmitted to the controller according to the offline designed routing and scheduling strategy. This process corresponds to the delay τ tc The controller calculates the control input u based on the full loop information age k , and transmit it to the downlink time-sensitive network. k In the downlink time-sensitive network, the routing and scheduling strategy designed offline is transmitted to the executor. This process corresponds to the delay τ ca Total delay τ = τ tc +τ ca The operation of the final actuator causes the controlled object to update its state and start the update process at the next moment.

[0106] In the first step, the time-sensitive network is designed based on the IEEE 802.1Qbv standard. In the offline case, a gated queue for switch transmission is formed by designing a deterministic transmission constraint set (including topology constraints, non-contention constraints, transposition constraints, non-loop constraints and routing constraints, etc.). During actual operation, each transmission port of the switch is periodically opened and closed according to the time specified by the gated queue designed offline to ensure the determinism of data transmission. Compared with the traditional time-sensitive network design, this technical solution includes an event trigger based on the age of full-loop information, which can coordinate the gated queue of the time-sensitive network, avoid the opening and closing of the switch port when there is no event trigger, and reduce the transmission cost.

[0107] In the second step, the control stability condition is a constraint on the average full-loop information age, specifically:

[0108]

[0109] in, is the average full-loop information age, ζ is the maximum value of the matrix trace, ∈ is the system dimension, σ is the noise variance, x0 is the initial state, is the weight coefficient matrix; when the average full-loop information age is less than the above expression, the system is stable.

[0110] In the third step, the event trigger based on the full loop information age is based on Δ + The value of is used to trigger the operation, specifically:

[0111]

[0112] Where ξ is the cost of opening and closing the gate of the time-sensitive switch once, It is the total number of times that all switch gates are opened and closed during the transmission of data in a time-sensitive network. in is the system matrix, ω is the environmental noise; Θ k+τ =L k+τ Λ k+τ L k+τ is the solution-dependent variable expression of the Riccati equation; when Δ + When the value of is greater than 0, the trigger occurs; when Δ + When the value is less than 0, the trigger does not occur.

[0113] In the third step, the full-loop information age is defined as

[0114] Δ k+τ =(1-δ k )(Δ k+τ-1 +1)+δ k τ

[0115] Among them, δ k is the event trigger variable, and τ is the total network transmission delay. If the event trigger occurs at time k, the value of the full loop information age is the total loop delay, that is, Δ k+τ =τ; if the event trigger does not occur at time k, the value of the full loop information age increases by 1 compared to its value at the previous time, that is, Δ k+τ =Δ k+τ-1 +1.

[0116] The δ k The value of Δ + Decide when Δ + When the value of is greater than 0, the trigger occurs, δ k =1; when Δ + When the value is less than 0, the trigger does not occur. k =0.

[0117] In the fourth step, the control strategy based on the full loop information age uses the LQR algorithm and defines new state information and input, which can make the system quickly stable while eliminating the influence of input delay, further ensuring the stability of the system. The new state information and input expressions are:

[0118]

[0119] The optimal control decision is:

[0120]

[0121] in, where Y k+1 is the solution of the following Riccati equation

[0122]

[0123]

[0124]

[0125] in,

[0126] In the fifth step, the system compares the total cost of the control transmission joint design based on the full-loop information age under each routing scheduling strategy, and takes the solution with the smallest total cost as the optimal solution in the feasible solution set, including the optimal routing scheduling strategy, the optimal control strategy and the optimal event triggering strategy. This solution can ensure the speed and stability of the control system, and reduce the amount of data transmission in the network as much as possible, reduce the network burden, and improve the overall performance of the industrial network system.

[0127] In this embodiment, the control transmission joint design architecture based on the full loop information age is as follows Figure 1 The full loop information age growth curve is shown as Figure 2 The control transmission joint design algorithm based on the full loop information age is shown in Figure 3 The physical topology connection diagram is shown in Figure 4 In this embodiment, the user needs to model the control and transmission system of the industrial cyber-physical system based on the full-loop information age, and obtain the optimal routing and scheduling strategy, the optimal control strategy, and the optimal event triggering strategy by solving the optimization problem. In this embodiment, the network topology is as follows Figure 2 As shown in Figure 1, it includes 10 control subsystems, a time-sensitive network (including 10 switches), and a controller. The overall implementation process is as follows: Figure 4 As shown, specifically:

[0128] Step 1: Input the network topology and the control periods and control system equations of the 10 control subsystems.

[0129] Step 2: According to the input topology and the control cycle of the data flow, a deterministic transmission constraint set in the time-sensitive network is established, and then M feasible routing and scheduling strategies are obtained through the optimization solution of Satisfiability Modulo Theories (SMT).

[0130] Step 3: Initialize n1, n2, k, which are the feasible routing and scheduling solution number, subsystem number, and system running time number respectively.

[0131] Step 4: Determine whether k is less than T-τ. If so, jump to step 5; otherwise, jump to step 9.

[0132] Step 5: Calculation And judge Is it greater than 0? If so, jump to step 6; otherwise, jump to step 8.

[0133] Step 6: Trigger occurs, record the trigger plan make Skip to step 8.

[0134] Step 7: If the trigger does not occur, record the trigger plan make

[0135] Step 8: Calculation Record Let temp=k+1, k=temp, and jump to step 4.

[0136] Step 9: Determine whether n2 is less than N. If so, jump to step 9; otherwise, jump to step 11.

[0137] Step 10: Calculation

[0138]

[0139]

[0140]

[0141] Skip to step 4.

[0142] Step 11: Determine whether n1 is less than M. If so, jump to step 12; otherwise, jump to step 18.

[0143] Step 12: Determine whether the control stability condition is met. If yes, jump to step 13; otherwise, jump to step 17.

[0144] Step 13: Calculation

[0145]

[0146] Step 14: Judgement Is it less than Ω? min If yes, go to step 15; otherwise, go to step 16.

[0147] Step 15: Let min=n1.

[0148] Step 16: Let temp=n1+1, n1=temp, n2=1, k=0, and jump to step 4.

[0149] Step 17: Order Skip to step 16.

[0150] Step 18: Let u * =u min ,γ * =γ min ,S * =S min , output u * ,γ * ,S * , the process ends. So far, the joint design problem of the control system and transmission system based on full-loop AoI has been solved.

[0151] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A control transmission joint design method based on full-loop information age, characterized in that: The following steps are involved: Step 1: For the industrial cyber-physical system, for a given network topology and the control cycle of each subsystem data flow, generate a feasible deterministic routing and scheduling scheme S; assume that there are M schemes; initialize the scheduling scheme counting variable n1; the industrial cyber-physical system contains N subsystems in total, initialize the subsystem counting variable n2; the design time of the industrial cyber-physical system is T, and initialize the design initial time k of subsystem k2; Step 2: If ,in is the total delay, jump to step 3; If and n2 < N, then let temp = n2 + 1, n2 = temp, k = 0, and return to step 2; If and n2 ≥ N and n1 < M, then determine whether the age of information of the full loop corresponding to the n1 scheduling scheme satisfies the control stability condition. If so, calculate the corresponding transmission cost, control cost, and total cost under the n1 scheduling scheme: And let temo=n1+1,n1=temp,n2=1,k=0, return to step 2; otherwise, let temp=n1+1,n1=temo,n2=1,k=0, return to step 2; like If n2≥N and n1≥M, then jump to step 5; Step 3: For the n1th subsystem, for the n2th delay set, calculate the corresponding time k judge Is it less than 0? If so, If not, then Step 4: According to and The value result is based on the full loop information age Δ k+τ Calculate the optimal control solution Let temp = k + 1, k = temp, and return to step 2; Step 5: For n1∈[1,M], compare the corresponding Select the smallest As the optimization result, it is recorded as the total cost of optimal control transmission Ω * ; The corresponding routing scheduling scheme S, control strategy c, and event triggering scheme γ are taken as the control transmission joint design results, denoted as S * 、c * , γ * ; In the step 1, the industrial cyber-physical system includes a controlled object, a sensor, an actuator, an event trigger based on the full-loop information age, a controller based on the full-loop information age, and a time-sensitive network; The state information x of the controlled object k The sensor collects and transmits it to the event trigger, which determines the current x based on the age of the full loop information. k Whether to trigger; if triggered, x k Transmitted to the uplink time-sensitive network; if not triggered, wait for receiving the next moment x k+1 , and judge whether it is triggered or not; x k After entering the uplink time-sensitive network, it is transmitted to the controller according to the routing and scheduling strategy designed offline. The process corresponds to the delay ; The controller calculates the control input u based on the full loop information age k , and transmitted to the downlink time-sensitive network; u k In the downlink time-sensitive network, the routing and scheduling strategies designed offline are transmitted to the executor, and the corresponding delay of the process Total delay ; The operation of the final actuator causes the controlled object to update its state and start the update process at the next moment; In step 3, the event trigger based on the full loop information age is based on Δ + The value of is used to trigger the operation, specifically: Where ξ is the cost of opening and closing the gate of the time-sensitive switch once, It is the total number of times that all switch gates are opened and closed during the transmission of data in a time-sensitive network. in is the system matrix, ω is the environmental noise; is the solution-dependent variable expression of the Riccati equation; when Δ + When the value of is greater than 0, the trigger occurs; when Δ + When the value is less than 0, the trigger does not occur; In step 3, the definition of the full loop information age is Among them, δ k For event trigger variables, is the total network transmission delay; if the event trigger occurs at time k, the value of the full loop information age is the total loop delay, that is, ; If the event trigger does not occur at time k, the value of the full loop information age increases by 1 compared to its value at the previous time, that is, ; δ k The value of Δ + Decide when Δ + When the value of is greater than 0, the trigger occurs, δ k =1; when Δ + When the value is less than 0, the trigger does not occur. k =0.

2. The control transmission joint design method based on full-loop information age as claimed in claim 1 is characterized in that: In the step 1, the time-sensitive network is designed based on the IEEE 802.1Qbv standard.

3. The control transmission joint design method based on full loop information age as claimed in claim 1, characterized in that: In step 1, in an offline state, a gated queue for switch transmission is formed by designing a deterministic transmission constraint set; in actual operation, each transmission port of the switch is periodically opened and closed according to the time specified by the gated queue designed offline to ensure the determinism of data transmission.

4. The control transmission joint design method based on full loop information age as claimed in claim 3 is characterized in that: The constraint set includes topology constraints, contention-free constraints, transposition constraints, loop-free constraints and routing constraints.

5. The control transmission joint design method based on full loop information age as claimed in claim 1, characterized in that: In the step 2, the control stability condition is a constraint on the average full-loop information age. When the average full-loop information age is less than the stability threshold, the system is stable.

6. The control transmission joint design method based on full loop information age as claimed in claim 1, characterized in that: In step 4, the control strategy based on the full-loop information age utilizes the LQR algorithm and defines new state information and input, thereby obtaining the optimal control decision.

7. The control transmission joint design method based on full loop information age as claimed in claim 1, characterized in that: In step 5, the system compares the total costs corresponding to the control transmission joint design based on the full-loop information age under each routing scheduling strategy, and takes the solution with the smallest total cost as the optimal solution in the feasible solution set, including the optimal routing scheduling strategy, the optimal control strategy and the optimal event triggering strategy.

Images