Method, device and computer readable medium for constructing a four-tank process system control model

Abstract

The application discloses a kind of four-tank process system control model construction method, device, equipment and computer readable medium, belong to automation control technical field.The method includes: multiple modal four-tank process system is modeled as semi Markov jump neural network model;Adaptive trigger protocol is used to establish the data sampling trigger condition of neural network model;The feedback controller of non-periodic denial-of-service attack signal under data sampling trigger condition is constructed, and the closed-loop model of neural network model is obtained according to feedback controller calculation;The sufficient condition for ensuring the exponential synchronization of closed-loop model and the controller gain solvable of feedback controller is calculated, and the control parameter of feedback controller is obtained.The technical scheme disclosed by the application can accurately describe the nonlinear dynamic characteristics of multiple modal four-tank process system under denial-of-service attack, and effectively suppress the influence of denial-of-service attack on multiple modal four-tank process system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automation control technology, and in particular, to a method and device for constructing a control model of a four-tank process system, and a computer readable medium. BACKGROUND

[0002] In the field of process control engineering, a four-tank process system is a system that simulates various complex controls in industrial production processes. The control of the four-tank process system not only integrates various technologies such as automation instruments, automatic control, and communication technology, but also includes controlled parameters such as liquid level, flow rate, and temperature. Currently, with the continuous expansion of production scale and the gradual improvement of product quality requirements, the liquid level control task of the four-tank process system is facing challenges such as high nonlinearity, large inertia, strong coupling, and large time delay, which will affect the stability of the control system. In addition, the four-tank process system is greatly affected by parameter and structure mutations, making the control task more complex, and accurate control of multiple variables is also very difficult. Therefore, it is an urgent problem to be solved to study the event-triggered control method of the multi-modal four-tank process system under denial-of-service attacks and to establish a controller to enable the multi-modal four-tank process system to operate stably under denial-of-service attacks. SUMMARY

[0003] To solve at least one of the technical problems existing in the prior art, embodiments of the present application provide a method and device for constructing a control model of a four-tank process system, and a computer readable medium. The technical solution is as follows:

[0004] In a first aspect, a method for constructing a control model of a four-tank process system is provided, the method comprising:

[0005] modeling a multi-modal four-tank process system as a semi-Markov jump neural network model, the neural network model comprising: a master system model and a slave system model;

[0006] establishing a data sampling trigger condition of the neural network model using an adaptive trigger protocol;

[0007] constructing a feedback controller of aperiodic denial-of-service attack signals under the data sampling trigger condition, and calculating a closed-loop model of the neural network model according to the feedback controller;

[0008] calculating sufficient conditions for ensuring exponential synchronization of the closed-loop model and solvability of the controller gain of the feedback controller, and obtaining control parameters of the feedback controller.

[0009] Further, the master system model comprises:

[0010]

[0011]

[0012] wherein, is a state with n neurons,

[0013] A(a(t)) is a positive definite matrix,

[0014] C(a(t)) and D(a(t)) are connection weight matrix and time-delay connection weight matrix, respectively,

[0015] E(t) is an external input,

[0016] σ(t) represents a time delay, and 0≤σ(t)≤σ M ,

[0017] represents an activation function of a neuron, and satisfies

[0018]

[0019] wherein g i (0) = 0, and are known scalars,

[0020] {a(t), t≥0} is a semi-Markov process, taking values in For a(t) = μ, its transition rate is

[0021]

[0022] wherein d≥0 represents a residence time, and is a transition rate from μ to , It is generally assumed that the transition rate is bounded, satisfying wherein and are known real scalars, defined as wherein and

[0023] Further, the system model comprises:

[0024]

[0025]

[0026] wherein, represents a state,

[0027] is a control input,

[0028] For the error vector The error system is described as

[0029]

[0030]

[0031] Wherein g(ε(t)) satisfies Wherein

[0032] Further, the data sampling trigger condition for establishing the neural network model using the adaptive trigger protocol comprises:

[0033]

[0034] Wherein, wherein Ξ μ > 0, Indicates the error state between the current data and the latest data,

[0035] And Respectively represent the sampling period and the latest trigger time,

[0036] Indicates the latest transmission data, The current sampling data.

[0037] Further, the feedback controller for constructing the non-periodic denial of service attack signal under the data sampling trigger condition comprises:

[0038]

[0039] Wherein, K μ The controller gain,

[0040] Wherein,

[0041] The number of trigger times during the nth denial of service attack, and

[0042]

[0043] Further, the closed-loop model of the neural network model is calculated according to the feedback controller, comprising:

[0044]

[0045]

[0046] wherein,

[0047] Further, the calculation guarantees a sufficient condition of exponential synchronization of the closed-loop model and controller gain solvability of the feedback controller, and obtains control parameters of the feedback controller, comprising:

[0048] By using random Lyapunov stability theory and integral inequality, a sufficient condition of exponential synchronization of semi-Markov jump neural network and controller gain solvability is calculated and obtained.

[0049] In a second aspect, a device for constructing a four-tank process system control model is provided, and the device comprises:

[0050] A modeling module is configured to model a multi-modal four-tank process system as a semi-Markov jump neural network model, wherein the neural network model comprises a master system model and a slave system model.

[0051] A sampling condition construction module is configured to establish a data sampling trigger condition of the neural network model by using an adaptive trigger protocol.

[0052] A closed-loop model construction module is configured to construct a feedback controller of a non-periodic denial-of-service attack signal under the data sampling trigger condition, and to calculate a closed-loop model of the neural network model according to the feedback controller.

[0053] A control parameter calculation module is configured to calculate a sufficient condition of exponential synchronization of the closed-loop model and controller gain solvability of the feedback controller, and to obtain control parameters of the feedback controller.

[0054] Further, the master system model constructed by the modeling module comprises:

[0055]

[0056]

[0057] wherein, is a state with n neurons,

[0058] E(t) is an external input,

[0059] σ(t) represents a time delay, and 0≤σ(t)≤σ M ,

[0060] represents an activation function of a neuron, and satisfies

[0061]

[0062] where g i (0) = 0, and are known scalars,

[0063] {a(t), t > 0} is a semi-Markov process taking values in For a(t) = μ, the transition rate is

[0064]

[0065] where d > 0 represents the residence time, and is the transition rate from μ to It is generally assumed that the transition rate is bounded, satisfying where and are known real scalars, defined by where and

[0066] Further, the modeling module constructs a system model from, including:

[0067]

[0068]

[0069] where, denotes the state,

[0070] is the control input,

[0071] For the error vector the error system is described as

[0072]

[0073]

[0074] where g(e(t)) satisfies where

[0075] Further, the sampling condition construction module constructs a data sampling trigger condition, including:

[0076]

[0077] where, where Ξ μ > 0, ​Indicates the error status between the current data and the latest data,

[0078] and Represented as sampling period and latest triggering time respectively,

[0079] Indicates the latest transmission data, The current sampling data.

[0080] In one embodiment, the feedback controller constructed by the closed-loop model construction module 803 includes:

[0081]

[0082] Among them, K μ is the controller gain,

[0083] in,

[0084] is the number of trigger moments during the nth DoS attack, and

[0085]

[0086] Furthermore, the closed-loop model building module is also used to:

[0087] The closed-loop model of the neural network model is obtained by calculation according to the feedback controller, including:

[0088]

[0089]

[0090] in,

[0091] Furthermore, the control parameter calculation module is specifically used to:

[0092] By using the stochastic Lyapunov stability theory and integral inequalities, sufficient conditions are obtained to ensure the exponential synchronization of semi-Markov jump neural networks and the solvability of controller gains.

[0093] According to a third aspect, an electronic device is provided, including:

[0094] one or more processors; and

[0095] A memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute any method described in the first aspect.

[0096] In a fourth aspect, a computer readable medium is provided, and a computer program is stored on the computer readable medium, wherein the program, when executed by a processor, implements the method according to any one of the first aspect.

[0097] The technical scheme provided by the embodiment of the application has the beneficial effects that:

[0098] 1. The technical scheme disclosed by the embodiment of the application uses a semi-Markov jump neural network model to describe a multi-modal four-tank process system model under a denial-of-service attack, and can better describe the dynamic characteristics thereof.

[0099] 2. The technical scheme disclosed by the embodiment of the application constructs a feedback control law based on an adaptive event triggering mechanism and basic characteristics of a denial-of-service attack, and ensures exponential synchronization of a closed-loop system.

[0100] 3. The technical scheme disclosed by the embodiment of the application can accurately describe nonlinear dynamic characteristics of a multi-modal four-tank process system under a denial-of-service attack, effectively suppress the influence of the denial-of-service attack on the multi-modal four-tank process system, solve the event-triggered control problem of the multi-modal four-tank process system, and improve the performance of the multi-modal four-tank process system. BRIEF DESCRIPTION OF DRAWINGS

[0101] In order to more clearly illustrate the technical scheme in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0102] Figure 1 is a flow chart of a method for constructing a four-tank process system control model provided by the embodiment of the application;

[0103] Figure 2 is a multi-modal four-tank process system diagram provided by the embodiment of the application;

[0104] Figure 3 is an error state trajectory simulation diagram of a multi-modal four-tank process system provided by the embodiment of the application;

[0105] Figure 4 is a control input simulation diagram of a multi-modal four-tank process system provided by the embodiment of the application;

[0106] Figure 5 is a denial-of-service attack signal simulation diagram provided by the embodiment of the application;

[0107] Figure 6 is a release interval and time simulation diagram under adaptive event triggering provided by the embodiment of the application;

[0108] Figure 7 is an adaptive threshold simulation diagram provided by an embodiment of the present application;

[0109] Figure 8 is a structural schematic diagram of a four-tank process system control model construction device provided by an embodiment of the present application;

[0110] Figure 9 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0111] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0112] The existing event-triggered control method of a multi-modal four-tank process system is difficult to adapt to stable operation under a denial-of-service attack. In order to solve the problems in the prior art, an embodiment of the present application provides a four-tank process system control model construction method, device, equipment and computer readable medium, and the specific technical solutions are as follows:

[0113] S1, model a multi-modal four-tank process system as a semi-Markov jump neural network model, the neural network model comprising: a master system model and a slave system model.

[0114] The above multi-modal four-tank process system model is modeled as a delayed semi-Markov jump neural network system as follows:

[0115]

[0116]

[0117] wherein is a state with n neurons. A(α(t)) is a positive definite matrix, C(α(t)) and D(α(t)) are connection weight matrices and time delay connection weight matrices, respectively. E(t) is an external input. σ(t) represents a time delay, and 0≤σ(t)≤σ M , represents an activation function of a neuron, and satisfies

[0118]

[0119] wherein g i (0)=0, and are known scalars.

[0120] {a(t), t≥0} is a semi-Markov process taking values in For a(t) = μ, the transition rate is

[0121]

[0122] where d≥0 represents the residence time, and is the transition rate from μ to It is generally assumed that the transition rate is bounded, satisfying where and are known real scalars. Define where and

[0123] In addition, the slave system model is described as follows

[0124]

[0125]

[0126] where denotes the state. is the control input, i.e., the feedback controller.

[0127] Define the error vector Then the error system of the master and slave system models can be described as

[0128]

[0129]

[0130] where g(ε(t)) satisfies where

[0131] S2, a data sampling trigger condition for establishing a neural network model using an adaptive event trigger protocol.

[0132] The above, considering the limitation of network bandwidth, in order to make the most effective use of network bandwidth resources, reduce unnecessary waste, adaptive event trigger protocol is adopted to save network bandwidth resources. Define as a periodic sampling sequence. and ​respectively, represent the sampling period and the latest triggering time. The next triggering time is

[0133]

[0134] where Ξ μ > 0, is the current sampling data, represents the latest transmission data, represents the error state between the current data and the latest data. In the network channel, the delay is inevitable in the process of signal transmission. Assuming the delay q ∈ {0, 1, 2,...}, the time t of the transmitted data reaching the controller satisfies

[0135] The adaptive threshold parameter τ(t) is a threshold variable satisfying the following adaptive law

[0136]

[0137] where 0 < τ(t) ≤ 1, and π > 0 is used to adjust the convergence of τ(t). Therefore, the triggering condition of the adaptive event-triggered protocol depending on the state error can be expressed as

[0138]

[0139] S3, construct a feedback controller of the non-periodic denial of service attack signal under the data sampling triggering condition, and calculate the closed-loop model of the neural network model according to the feedback controller.

[0140] As mentioned above, due to the opening of the network transmission channel, data is easy to be attacked by malicious network attacks when transmitted in the network. Considering the non-periodic denial of service attack, the data transmission is affected by consuming network communication resources. The data triggered and transmitted by the adaptive event-triggered protocol is subjected to non-periodic denial of service attack in the network channel, and the variable λ(t) is used to represent the non-periodic denial of service attack signal:

[0141]

[0142] where λ(t) = 1 represents that the system is not subjected to denial of service attack. λ(t) = 0 represents that the system is subjected to denial of service attack. w n and w n + r n represent the start time and end time of the n th sleep of the denial of service attack, w n+1 - w n - r n represent the duration of the n th attack. The interval of the denial of service attack can be represented as T 1,n = [wn ,w n +r n ) and T 2,n = [w n +r n ,w n+1 ).

[0143] Assumption 1: Assume that during the interval of denial-of-service attack there exists: where w m and w M denote the lower bound of the sleep period r n and the upper bound of the attack period w n+1 -w n -r n , respectively.

[0144] Assumption 2: For the interval [t0, t), if given f > 0 and then the sequence of denial-of-service attack satisfies where Ω(t) denotes the number of denial-of-service attack sleep / activation transitions.

[0145] S4, calculate the sufficient condition of the controller gain that guarantees the closed-loop model exponential synchronization and the feedback controller, and obtain the control parameters of the feedback controller.

[0146] In the above, in combination with the adaptive event-triggered protocol and the aperiodic denial-of-service attack on the system, the feedback controller can be designed as follows:

[0147]

[0148] where K μ is the controller gain.

[0149] However, in the case of adaptive event-triggered protocol and aperiodic denial-of-service attack, the data triggered and transmitted can only continue to run in T 1,n , while in T 2,n be interrupted. Obviously, the condition of adaptive event-triggered protocol is not applicable to T 2,n . Therefore, the condition of adaptive event-triggered protocol is updated to adapt to the existence of aperiodic denial-of-service attack, and the triggering time is

[0150]

[0151] where is the number of triggering times during the nth denial-of-service attack, and

[0152] Define then the time interval Y q,nDivided into where q ∈ γ(n), l ∈ {1,2,...,ρ q,n} and

[0153] Let then

[0154] Definition

[0155]

[0156]

[0157] where b = 1,2,...,ρ q,n +1, t ∈ Y q,n ∩ T 1,n .

[0158] Thus, the sampled data transmitted to the communication network is

[0159]

[0160] where κ q,n (t) and τ(t) satisfy

[0161]

[0162]

[0163] In combination with the designed controller, the closed-loop system can be obtained in the form of

[0164]

[0165]

[0166] where

[0167] In one embodiment, further, the control model constructed in the above steps S1-S4 is subjected to synchronization analysis of the closed-loop active system.

[0168] For a given scalar w M ,w m ,f,k,η,π>0,σ M >0,γ M >0, if there exists a scalar ξ i >0,c1>0,c2>0,p μ , a symmetric positive definite matrix P iμ ,Q 1i ,Q2i , 1i , 2i , 1i , 2i Λ is a matrix of appropriate dimension μ , 1i , 2i J1, J2 are diagonal matrices, and υ = 1, 2, i = 1, 2, such that the following inequalities are satisfied

[0169] Φ1 + W < 0,

[0170] Φ2 < 0,

[0171]

[0172]

[0173] U 1υ ≤ c 3-υ U 1(3-υ) , 2υ ≤ c 3-υ U 2(3-υ) ,

[0174] Q 1υ ≤ c 3-υ Q 1(3-υ) , 2υ ≤ c 3-υ Q 2(3-υ) ,

[0175] R 1υ ≤ c 3-υ R 1(3-υ) , 2υ ≤ c 3-υ R 2(3-υ) ,

[0176]

[0177] where

[0178]

[0179]

[0180] θ1= [ω1 ω4], θ2= [ω2 ω5], θ3= [ω1 ω2 ω3], θ4= [ω1 ω6 ω7],

[0181]

[0182]

[0183]

[0184] φ(t) = [ε T (t) ε T (t-σ(t)) ε T (t-σ M ) g T (ε(t)) g T (ε(t-σ(t)))

[0185]

[0186] where ω2φ(t) = ε(t-σ(t)), the system is exponentially synchronized in the mean square sense.

[0187] In order to make the four-tank process system control model construction method provided by the embodiments of the present application more detailed, the technical solutions disclosed by the present application are exemplarily demonstrated below in combination with specific application scenarios.

[0188] The original process principle diagram of the multi-modal four-tank process system model is shown in Figure 2 The control target is to control the liquid levels of two low tanks using two pumps. Therefore, the input voltages of Tank1 and Tank2 are Tank1 and Tank2, and the output is the measured water levels h1 and h2.

[0189] S1, model the multi-modal four-tank process system as a semi-Markov jump neural network model, and the neural network model comprises: a master system model and a slave system model.

[0190] By using Bernoulli's law, fluid mechanics principles and mass conservation principles, the dynamic mathematical model of the four-tank water tank system can be described by a state equation as follows:

[0191]

[0192]

[0193]

[0194]

[0195] where g is the acceleration of gravity, h i (i = 1, 2, 3, 4) is the water level in tank i, w1 and w2 are manipulated inputs (voltage applied to the pump), the water output is k1w1 and k2w2, respectively, S i is the cross section of tank i, m i is a parameter related to the pressure difference of the tank output flow. The proportion of water that is diverted to tank 1 instead of tank 3 is r1. Similarly,

[0196] Similarly, r2 is the corresponding ratio of tank 2 and tank 4. In this model, the transmission delay d in the circuit is taken into account. i (i=1,...,6). If it is completely symmetrical, then d1=d2, d3=d4 and d5=d6. Let the equilibrium point of the system be h 10 ,h 20 ,h 30 ,h 40 ,w 10 ,w 20 , the equation is linearized at the equilibrium point, and the variable is introduced The following kinetic model is obtained:

[0197]

[0198] in

[0199]

[0200]

[0201] The model parameters are selected as W1=W2=W3=W4=9.2cm, h max =25cm, g=981cm / s 2 , k1=7.3972,k2=6.9253. Then we have

[0202]

[0203]

[0204]

[0205] Due to the mutation of some parameters, r1 and r2 choose two different modes, namely r 11 =0.333, r 12 =0.450, r 21 =0.307, r 22 =0.380. The jump between different ratios follows a semi-Markov process {α(t),t≥0}, with a value of parameter and Related to the semi-Markov process α(t), they are expressed as and

[0206] For the transfer rate Given δ 11 =-1,δ 12 =1,δ21 = 0.8, δ 22 = -0.8, Ψ 11 = Ψ 12 = 0.05, Ψ 21 = Ψ 22 = 0.04.

[0207] Let d1 = d5 = 0, d3 = σ(t) = 0.18 + 0.18sin(t), the water supply is represented as j = 1,..., 4. The multi-modal four-tank process system model can be described as:

[0208]

[0209] where E(t) = [0 0 0 0].

[0210] S2, the data sampling trigger condition of the neural network model established by the adaptive trigger protocol.

[0211] An adaptive event-triggered protocol is used to save network bandwidth resources, and is defined as a periodic sampling sequence. and are represented as the sampling period and the latest trigger time, respectively. The next trigger time is

[0212]

[0213] where Ξ μ > 0, is the current sampling data, represents the latest transmission data, represents the error state between the current data and the latest data. In the network channel, time delay inevitably occurs in the signal transmission process. Assuming that the delay q ∈ {0, 1, 2,...}, then the time t when the transmitted data arrives at the controller satisfies

[0214] The adaptive threshold parameter τ(t) is a threshold variable that satisfies the following adaptive law:

[0215]

[0216] where 0 < τ(t) ≤ 1, and π = 6 is selected to adjust the convergence of τ(t). Therefore, the trigger condition of the adaptive event-triggered protocol depending on the state error can be represented as

[0217]

[0218] S3, construct a feedback controller of non-periodic denial-of-service attack signal under the condition of data sampling trigger, and obtain the closed-loop model of the neural network model according to the feedback controller.

[0219] The variable λ(t) is used to represent the non-periodic denial-of-service attack signal

[0220]

[0221] Wherein λ(t) = 1 indicates that the system is not attacked by denial-of-service attack. λ(t) = 0 indicates that the system is attacked by denial-of-service attack. n And w n +r n represents the start time and end time of the n th sleep of denial-of-service attack, w n+1 -w n -r n represents the duration of the n th attack. The interval of denial-of-service attack can be represented as T 1,n =[w n ,w n +r n ) and T 2,n =[w n +r n ,w n+1 ).

[0222] Hypothesis 1: it is assumed that during the interval of denial-of-service attack, there is: Wherein w m and w M represent the lower bound of sleep period r n and the upper bound of attack period w n+1 -w n -r n .

[0223] Hypothesis 2: for the interval [t0, t), if given f > 0 and The sequence of denial-of-service attacks satisfies Wherein Ω(t) represents the number of denial-of-service attack sleep / activation transitions.

[0224] Select w M = 0.6, w m = 1.38, f = 2, k = 4.

[0225] S4, calculate the sufficient condition of ensuring the exponential synchronization of the closed-loop model and the solvability of the controller gain of the feedback controller, and obtain the control parameters of the feedback controller.

[0226] Combined with the adaptive event-triggered protocol and the non-periodic denial-of-service attack on the system, the feedback controller can be designed as follows:

[0227]

[0228] where K μ is the controller gain.

[0229] However, in the case of adaptive event-triggered protocol and aperiodic denial-of-service attack, the triggered and transmitted data can only continue running in T 1,n , while being interrupted in T 2,n . Obviously, the condition of adaptive event-triggered protocol is not applicable to T 2,n . Therefore, the condition of adaptive event-triggered protocol is updated to adapt to the existence of aperiodic denial-of-service attack, and the triggering time is

[0230]

[0231] where is the number of triggering time during the nth denial-of-service attack, and

[0232] Define then the time interval Y q,n can be divided into where q∈γ(n), l∈{1,2,...,ρ q,n} and

[0233] Let then

[0234] Define

[0235]

[0236]

[0237] where b=1,2,...,ρ q,n +1, t∈Y q,n ∩T 1,n .

[0238] Thus, the sampled data transmitted to the communication network is

[0239]

[0240] where κ q,n (t) and τ(t) satisfy

[0241]

[0242]

[0243] With the designed controller, the closed-loop system can be obtained as

[0244]

[0245]

[0246] where

[0247] Select σ M = 0.18, γ M = 0.04.

[0248] S5, the synchronization analysis of the closed-loop active system is carried out on the control model constructed above.

[0249] For given scalar w M , w m , f, k, η, π > 0, σ M > 0, γ M > 0, if there is a scalar ξ i > 0, c1 > 0, c2 > 0, p μ , a symmetric positive definite matrix P iμ , Q 1i , Q 2i , U 1i , U 2i , R 1i , R 2i , a matrix Λ of appropriate dimension μ , T 1i , T 2i , diagonal matrix J1, J2, matrix υ = 1, 2, i = 1, 2, so that the following inequalities are established

[0250] Φ1 + W < 0,

[0251] Φ2 < 0,

[0252]

[0253]

[0254] U 1υ ≤ c 3-υ U 1(3-υ) , U 2υ ≤ c 3-υ U 2(3-υ) ,

[0255] Q 1υ ≤ c 3-υ Q 1(3-υ) , Q 2υ≤ c 3-υ Q 2(3-υ) ,

[0256] R 1υ ≤ c 3-υ R 1(3-υ) ,R 2υ ≤ c 3-υ R 2(3-υ) ,

[0257]

[0258] where

[0259]

[0260]

[0261] θ1=[ω1 ω4], θ2=[ω2 ω5], θ3=[ω1 ω2 ω3], θ4=[ω1 ω6 ω7],

[0262]

[0263]

[0264]

[0265]

[0266] where denotes the block input matrix, for example, ω2φ(t)=ε(t-σ(t)), then the system is exponentially synchronized in the mean square sense.

[0267] Selecting η=0.2, ξ1=0.015, ξ2=0.02, c1=c2=1.02, p1=p2=1.0, we can obtain:

[0268]

[0269] In order to clearly show the event-triggered control method proposed in the patent, part of the data trajectory is plotted in Figures 3-7 , wherein the initial state is taken as:

[0270] z(0)=[-0.4-0.50.30.7] T ,

[0271] Figure 3 The error state trajectory is embodied. It can be seen that the error value converges to zero after about 7s, which indicates that the synchronization between the system and the main system has been achieved. Figure 4 The control input is described, and the control input converges to the origin under the denial of service attack.Figure 5 The evolution of denial of service attack is drawn, and the non-periodic characteristics of the attack are embodied. Figure 6 The release interval and time under adaptive event triggering are described, and the threshold changes as shown. Figure 7 According to the method, the influence of the denial of service attack on the multi-modal four-tank process system can be effectively inhibited. Figures 3-7 The event-triggered control problem of the multi-modal four-tank process system under the denial of service attack can be solved, and the performance of the multi-modal four-tank process system is improved.

[0272] Based on the control parameter calculation module disclosed in the embodiment of the application, as shown in the figure, the embodiment of the application further provides a construction device of a control model of a four-tank process system, comprising: Figure 8 The modeling module 801 is configured to model the multi-modal four-tank process system into a semi-Markov jump neural network model, and the neural network model comprises a master system model and a slave system model.

[0273] The sampling condition construction module 802 is configured to establish a data sampling triggering condition of the neural network model by using an adaptive triggering protocol.

[0274] The closed-loop model construction module 803 is configured to construct a feedback controller of a non-periodic denial of service attack signal under the data sampling triggering condition, and to calculate a closed-loop model of the neural network model according to the feedback controller.

[0275] The control parameter calculation module 804 is configured to calculate sufficient conditions for ensuring exponential synchronization of the closed-loop model and solvability of controller gains of the feedback controller, and to obtain control parameters of the feedback controller.

[0276] In one embodiment, the master system model constructed by the modeling module 801 comprises:

[0277]

[0278]

[0279]

[0280] wherein, is a state with n neurons,

[0281] E(t) is an external input,

[0282] σ(t) represents a time delay, and 0≤σ(t)≤σ M ,

[0283] represents an activation function of a neuron, and satisfies

[0284] ​

[0285] where g i (0) = 0, ι1≠ι2, and is a known scalar,

[0286] {α(t), t≥0} is a semi-Markov process taking values in For α(t) = μ, its transition rate is

[0287]

[0288] where d≥0 represents the residence time, and is the transition rate from μ to It is generally assumed that the transition rate is bounded, satisfying where and are known real scalars, defined by where and

[0289] In one embodiment, the modeling module 801 constructs a system model from, including:

[0290]

[0291]

[0292] where, denotes the state,

[0293] is the control input,

[0294] Define the error vector The error system is described as

[0295]

[0296]

[0297] where g(ε(t)) satisfies where

[0298] In one embodiment, the sampling condition construction module 802 constructs a data sampling trigger condition, including:

[0299] ​

[0300] wherein, wherein μ > 0, denotes the error state between the current data and the latest data,

[0301] and denote the sampling period and the latest triggering time, respectively,

[0302] denotes the latest transmission data, is the current sampling data.

[0303] In one embodiment, the closed-loop model construction module 803 constructs a feedback controller, comprising:

[0304]

[0305] wherein, K μ is a controller gain,

[0306] wherein,

[0307] is the number of triggering times during the nth denial-of-service attack, and

[0308]

[0309] In one embodiment, the closed-loop model construction module 803 is further configured to:

[0310] obtain a closed-loop model of the neural network model according to the feedback controller, comprising:

[0311]

[0312]

[0313] wherein,

[0314] In one embodiment, the control parameter calculation module 804 is specifically configured to:

[0315] obtain a sufficient condition for ensuring the exponential synchronization of the semi-Markov jump neural network and the solvability of the controller gain by using the random Lyapunov stability theory and the integral inequality.

[0316] As described above, the control model construction device of the four-tank process system disclosed in the present invention executes the main steps of the control parameter calculation module method disclosed in the present invention through the modeling module, the sampling condition construction module, the feedback controller construction module, and the control parameter calculation module. The device can be set separately in a hardware device, or can be nested in a process control system.

[0317] In addition, an embodiment of the present invention further provides an electronic device, including:

[0318] one or more processors; and

[0319] A memory associated with the one or more processors, the memory being used to store program instructions, wherein when the program instructions are read and executed by the one or more processors, the method for constructing a control model of a four-tank process system disclosed in the above embodiment is executed.

[0320] Among them, such as Figure 9 As shown, computer device 12 is implemented as a general-purpose computing device. Components of computer device 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, and a bus 18 that connects various system components (including system memory 28 and processing unit 16). Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0321] The computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0322] System memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer device 12 can further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 can be provided for reading from and writing to non-removable, non-volatile magnetic media (not shown and typically called a "hard drive"). Although not shown, a magnetic disk drive can also be used for reading from and writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive can be used for reading from or writing to a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM or other optical media). In such instances, each can be connected to bus 18 by one or more data media interfaces. As will be further depicted and described below, memory 28 can include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the application.

[0323] Program / utility 40, having a set (at least one) of program modules 42, can be stored in, for example, memory 28 by way of example, and not limitation, as well as an operating system, one or more application programs, other program modules, and program data, each or some combination thereof, can include implementation of a networking environment. Program modules 42 generally carry out the functions and / or methodologies of embodiments of the application as described herein.

[0324] Computer device 12 can also communicate with one or more external devices 14 such as a keyboard, a pointing device, a display 24, etc.; one or more devices that enable a user to interact with computer device 12; and / or one or more devices that enable computer device 12 to communicate with one or more other computing devices. Such communication can be via input / output (I / O) interfaces 22. Additionally, in embodiments where computer device 12 is embedded in a mirror, display 24 is not present as a separate entity, but rather the display surface of display 24 is integrated with the mirror surface such that the display surface of display 24 is visually merged with the mirror surface when the display surface of display 24 is not displaying. Still yet, computer device 12 can communicate with one or more networks, such as a local area network (LAN), a wide area network (WAN), and / or the Internet, through network adapter 20. As depicted, network adapter 20 communicates with the other components of computer device 12 via bus 18. It should be appreciated that although not shown, other hardware and / or software modules could be used in conjunction with computer device 12. Such as, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0325] The processing unit 16 performs various function applications and data processing by running programs stored in the system memory 28.

[0326] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, the system or system embodiments are described simply because they are basically similar to the method embodiments. The relevant parts can be referred to the description of the method embodiments. The above-described system and system embodiments are merely illustrative. The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on multiple network units. Some or all of the modules can be selected to achieve the purpose of the embodiment according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0327] The above describes the technical solutions provided by the present application in detail. The principles and implementation modes of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method and core idea of the present application. For those skilled in the art, according to the idea of the present application, the specific implementation mode and application range can be changed. In summary, the content of the specification should not be understood as a limitation of the present application.

[0328] All the optional technical solutions described above can be combined to form optional embodiments of the present application, which will not be described here.

[0329] The above description is only the preferred embodiment of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for constructing a control model for a four-tank process system, characterized in that: include: The multimodal four-tank process system is modeled as a semi-Markov jump neural network model, wherein the neural network model includes: a master system model and a slave system model; The main system model is specifically calculated as follows: in, is a state with n neurons, A(α(t)) is a positive definite matrix, C(α(t)) and D(α(t)) are the connection weight matrix and the delay connection weight matrix respectively, E(t) is the external input, σ(t) represents the time delay, and 0≤σ(t)≤σ M , Represents the activation function of the neuron, and satisfies where g i (0)=0, ι1≠ι2, and is a known scalar, {α(t), t≥0} is a semi-Markov process. For α(t) = μ, its transfer rate is: in d≥0 represents the residence time, and From μ to The transfer rate, It is generally assumed that the transfer rate is bounded, satisfied in and is a known real scalar, and is defined as in and The specific calculation method of the slave system model is: in, Indicates status, is the control input, For the error vector The error system is described as in g(ε(t)) satisfies in ι≠0; Adopting an adaptive trigger protocol to establish a data sampling trigger condition for the neural network model; Constructing a feedback controller for the aperiodic denial of service attack signal under the data sampling trigger condition, and calculating and obtaining a closed-loop model of the neural network model based on the feedback controller; Sufficient conditions for ensuring exponential synchronization of the closed-loop model and solvability of the controller gain of the feedback controller are calculated to obtain control parameters of the feedback controller.

2. The method according to claim 1, wherein The data sampling triggering condition of the neural network model is established by adopting the adaptive triggering protocol, including: Among them, among them μ >0, Indicates the error status between the current data and the latest data, and Represented as sampling period and latest triggering time respectively, Indicates the latest transmission data, The current sampling data.

3. The method according to claim 1, wherein The feedback controller for constructing the non-periodic denial of service attack signal under the data sampling trigger condition includes: Among them, K μ is the controller gain, in, is the number of trigger moments during the nth DoS attack, and 4. The method according to claim 1, wherein The closed-loop model of the neural network model is obtained by calculating according to the feedback controller, comprising: where \(l=\max\{\sigma M ,\gamma M \}\).

5. The method according to claim 1, wherein The calculating of sufficient conditions to ensure exponential synchronization of the closed-loop model and solvability of the controller gain of the feedback controller, and obtaining control parameters of the feedback controller, includes: By using the stochastic Lyapunov stability theory and integral inequalities, sufficient conditions are obtained to ensure the exponential synchronization of semi-Markov jump neural networks and the solvability of controller gains.

6. An electronic device, characterized in that: include: one or more processors; as well as A memory associated with the one or more processors, the memory being used to store program instructions, wherein when the program instructions are read and executed by the one or more processors, the method according to any one of claims 1 to 5 is executed.

7. A computer-readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images