Distributed cooperative scheduling method and system based on demand driving in unreliable network

Abstract

The invention discloses a demand-driven distributed collaborative scheduling method and system in an unreliable network, and the method is characterized in that the method comprises the following steps: constructing a multi-agent model composed of a plurality of individual independent generator sets, and solving the scheduling parameters of a consistency error system through the collaborative error convergence of the multi-agent; describing a data updating demand by using a decision function so as to obtain a data interaction driving mechanism among multiple agents containing scheduling parameters; and judging whether a transmission attempt exists in the attack interval, and implementing multi-agent distributed cooperative scheduling for the attack transmission attempt according to a data interaction driving mechanism and a secure transmission scheduling protocol.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and more specifically, to a demand-driven distributed collaborative scheduling method and system in an unreliable network. Background Art

[0002] In recent years, with the vigorous development of renewable energy, the randomness of renewable energy generation such as wind power and photovoltaic power has made the smooth dispatch of the power system a hot issue. The main power plant generators in the current power system are approaching saturation, and it is difficult to increase the increment of thermal power units. In the short term, relatively stable clean energy such as hydropower and nuclear power can hardly make up for the current power shortage, especially in the summer when the air conditioning load surges and the winter when the unit power drops due to heating. In order to improve the flexible capacity and peak-shaving capacity of the power system, it is necessary to control the voltage and frequency of small generators in local power plants.

[0003] Generally, small-power generators have small unit capacity and are geographically dispersed, which makes it difficult to dispatch small-power generators at peak output. It is difficult for the dispatch center to give dispatch instructions for the output of individual units. However, distributed small-power generators can form a distributed network through networking, and the units within the network are connected to the common coupling point of the region. Modeling the networked units in the same region as a multi-agent network and using the distributed collaborative control dispatching scheme provides a feasible solution for peak output dispatching.

[0004] However, the network between nodes of distributed generator sets is vulnerable to various network interferences or malicious attacks due to its openness. In particular, when the network is unreliable, packet loss and communication interruption of data transmission, delay and rejection of dispatch signals by generator sets, and failure and over-limit risks of transmission lines will have a serious impact on the coordinated dispatch between distributed generator sets. In this case, there is a strong demand for analysis, synthesis and design methods to ensure the safety and reliability of power systems.

[0005] In the research related to power system, the refusal of load dispatch is usually simulated as an attack, and the load frequency under this attack is controlled, but the relevant research has not proposed corresponding methods from the perspective of transmission mechanism and elastic performance quantification. Generally, the existing research considers that the attack on the distributed system is synchronous, that is, it is assumed that the attack on all power transfer channels occurs and stops at the same time. In the research related to elastic control, the traditional communication DoS attack is still used, and the stability analysis of the system is given under the assumption of attack frequency and duration. However, the traditional control methods are all based on time-driven methods, that is, control is achieved according to a fixed sampling period. Although periodic sampling may be desirable from the perspective of analysis and design, it will lead to unnecessary waste of limited resources, especially in the process of networking and dispatching of small-power generators in local power plants. Due to the large number of units and scattered geographical distribution, the communication burden of the dispatching center is even greater. Under this premise, the distributed generators do not actively participate in the dispatching or accurately respond to the dispatching in real time, but only communicate when necessary. In this case, the traditional time-driven control algorithm is no longer applicable, and it is necessary to develop a new dispatching control algorithm that fully considers the networking factors to solve the problem of elastic collaborative dispatching of distributed generators under unreliable network communication.

[0006] In view of the above problems, there is an urgent need for a demand-driven distributed collaborative scheduling method and system in an unreliable network. Summary of the invention

[0007] In order to address the deficiencies in the prior art, the present invention provides a demand-driven distributed collaborative scheduling method and system in an unreliable network. The method constructs a multi-agent model and a collaborative error model to solve the scheduling parameters based on the converged error under the premise of fully covering the attack range, and implements scheduling based on the collaborative scheduling protocol under a known model to achieve short-term convergence of the system.

[0008] The present invention adopts the following technical solution.

[0009] The first aspect of the present invention relates to a demand-driven distributed collaborative scheduling method in an unreliable network, characterized in that the method includes the following steps: constructing a multi-agent model composed of multiple independent generator sets, and solving the scheduling parameters of the consistency error system through the collaborative error convergence of the multi-agent; using a decision function to describe the demand for data update to obtain a data interaction driving mechanism between the multi-agents containing the scheduling parameters; determining whether there is a transmission attempt within the attack interval, and implementing distributed collaborative scheduling of the multi-agents for the attack transmission attempt based on the data interaction driving mechanism and the secure transmission scheduling protocol.

[0010] Preferably, a multi-agent model consisting of a plurality of independent generator sets is constructed, including: the multi-agent model is:

[0011]

[0012] In the formula, x i (t) is the true output of the ith agent,

[0013] u i (t) is the scheduling input of the ith agent,

[0014] ω i (t) is the perturbation of the ith agent,

[0015] A, B and E are coefficient matrices respectively.

[0016] Preferably, solving the scheduling parameters of the consistency error system through the collaborative error convergence of the multi-agents includes: calculating the collaborative state when the multi-agents are coupled And analyze the coordination error ε corresponding to the coordination state i (t); Utilize the collaborative error of each intelligent agent, the adjacency error between multiple intelligent agents, and the collaborative disturbance of each intelligent agent to construct a consistent error system for the collaborative error; Select the Lyapunov function of the consistent error system, and solve the scheduling parameters involved in the consistent error system according to the attenuation rate of the Lyapunov function; The scheduling parameters include a scheduling gain matrix, a connectivity matrix, a convergence coefficient matrix, a phase interference coefficient matrix, a first weight, and a second weight.

[0017] Preferably, the scheduling parameters involved in the consistency error system are solved according to the attenuation rate of the Lyapunov function, including: limiting the value range of the attenuation rate of the Lyapunov function according to the preset equivalent attenuation rates of different transmission lines in the attack interval and the preset equivalent attenuation rates of different transmission lines in the non-attack interval.

[0018] Preferably, the equivalent attenuation rates of the different transmission lines in the attack interval and the equivalent attenuation rates of the different transmission lines in the non-attack interval are preset through an allowable attack intensity index; the allowable attack intensity index is related to the associated parameters of the duration of the attack on the multi-agent model collected historically within a preset time interval.

[0019] Preferably, a decision function is used to describe the need for data update to obtain a data interaction driving mechanism between multiple agents including scheduling parameters, including: the decision function is:

[0020]

[0021] The data interaction driving mechanism between the multiple agents is:

[0022]

[0023] In the formula, Υ i (t) is the dynamic auxiliary variable of the ith agent at time t,

[0024] e i (t) is the prediction error, that is,

[0025] Ω i is the convergence coefficient of the ith agent,

[0026] θ i is the interference interaction coefficient of the ith agent, and

[0027] and are the predicted outputs between the ith agent and the jth agent at time t,

[0028] a ij is the connectivity coefficient between the ith agent and the jth agent,

[0029] N is the total number of agents,

[0030] is the time when the i-th agent requests scheduling for the kth time.

[0031] Preferably, it is determined whether there is a transmission attempt in the attack interval, and distributed collaborative scheduling of multiple agents is implemented for the transmission attempt according to the data interaction driving mechanism and the secure transmission scheduling protocol, including: the time sequence of multiple attack interval transmission attempts between the i-th agent and the j-th agent is:

[0032]

[0033] In the formula, The set of allowed time for request scheduling of the i-th agent;

[0034] like Then respond to the current scheduling request and adjust the scheduling input of the i-th agent based on the controller gain;

[0035] like Then delay the current scheduling request to Δ i is the preset time interval.

[0036] The second aspect of the present invention relates to a demand-driven distributed collaborative scheduling system in an unreliable network; the system is implemented using a demand-driven distributed collaborative scheduling method in an unreliable network described in the first aspect of the present invention; wherein the system includes a construction module, a description module and a scheduling module; wherein the construction module is used to construct a multi-agent model composed of multiple independent individual generator sets, and solve the multi-agent model through the collaborative error convergence of the multi-agent; the description module is used to use a decision function to describe the demand for data update to obtain a data interaction driving mechanism between the multi-agents; the scheduling module is used to determine whether there is a transmission attempt within the attack interval, and implement distributed collaborative scheduling of the multi-agents for the transmission attempt according to the data interaction driving mechanism and the secure transmission scheduling protocol.

[0037] A third aspect of the present invention relates to a terminal, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method according to the first aspect of the present invention.

[0038] A fourth aspect of the present invention relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect of the present invention.

[0039] The beneficial effect of the present invention lies in that, compared with the prior art, the present invention is a demand-driven distributed collaborative scheduling method and system in an unreliable network. The method constructs a multi-agent model and a collaborative error model to solve the scheduling parameters based on the converged error under the premise of fully covering the attack range, implements scheduling based on the collaborative scheduling protocol under a known model, and achieves short-term convergence of the system.

[0040] The beneficial effects of the present invention also include:

[0041] 1. The method adjusts the convergence range of the collaborative error model according to the convergence speed of the communication network and transmission lines between multi-agents. Under the requirement that the convergence range is sufficient to cover the historical attack samples, it provides a reliable basis for achieving rapid convergence between multi-agents at the lowest cost.

[0042] 2. The method uniformly models random or malicious dispatch interruptions as independent denial of service attacks, that is, the blocking interference between the dispatch paths of distributed agents is time-varying and unknown. This method overcomes the defects of using random probability to describe distributed agents and improves the reliability of the power system analysis method.

[0043] 3. The method comprehensively considers the risks caused by load fluctuations, fluctuations of distributed generator sets, abnormalities of transmission lines, etc., realizes the voltage and frequency control of small generator sets of local power plants, improves the flexible capacity and peak-shaving ability of the power system, and effectively utilizes the capacity of small-power units of local power plants that are easily overlooked to make up for the shortage during peak electricity consumption.

[0044] 4. The method overcomes the problems of large number of units, dispersed geographical distribution, heavy burden on the dispatching center, etc., and solves the problem of elastic coordinated dispatching of distributed generators under unreliable networks with the minimum transmission construction cost. The method supports the adjustment or expansion of parameters such as redundant power of transmission lines, dispatchable capacity of generator sets, and fault tolerance time of transmission lines to obtain more reasonable convergence speed and error stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of a demand-driven distributed collaborative scheduling method in an unreliable network and a common coupling mode of a single intelligent agent in the system according to the present invention;

[0046] Figure 2 A schematic diagram of a demand-driven distributed collaborative scheduling method in an unreliable network and a time-varying system of a single intelligent agent in the system according to the present invention;

[0047] Figure 3 A schematic diagram of a multi-agent connection mode in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0048] Figure 4 A schematic diagram of independent denial of service attacks on different connected edges in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0049] Figure 5 A schematic diagram of a system operation trajectory monitored by a first agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0050] Figure 6 A schematic diagram of a system operation trajectory monitored by a second agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0051] Figure 7 A schematic diagram of a system operation trajectory monitored by a third agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0052] Figure 8A schematic diagram of a system operation trajectory monitored by a fourth agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0053] Fig. 9 A schematic diagram of input control signals of four agents in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0054] Fig.10 A discrete distribution diagram of scheduling requests of four intelligent agents in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention;

[0055] Fig.11 The figure is an evolution diagram of dynamic auxiliary variables of four intelligent agents in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solution and advantages of the present invention clearer and more accurate, the technical solution of the present invention is described in detail below through multiple specific implementation methods. The embodiments adopted by the present invention are only used to explain the present invention and are not used to limit the content of the present invention.

[0057] A first aspect of the present invention relates to a demand-driven distributed collaborative scheduling method in an unreliable network, the method comprising steps 1 to 3.

[0058] Step 1: construct a multi-agent model consisting of multiple independent generator sets, and solve the multi-agent model through collaborative error convergence of the multi-agents.

[0059] Figure 1 The present invention is a distributed collaborative scheduling method based on demand-driven in an unreliable network and a schematic diagram of a common coupling mode of a single intelligent agent in the system. Figure 1 As shown, a single distributed generator node connected to the common coupling point is extracted, the unknown load current is modeled as a disturbance, and the dynamic equation under the abc-framework is converted into a dq-model using Park transformation.

[0060] For each agent, the model is:

[0061] DGU i:

[0062] Where V i,dq ,I i,dq are the output voltage and current of the common coupling point of the intelligent body in the dq direction, and the corresponding dispatch output is expressed as V ti,dq ,I ti,dq, the corresponding disturbance output is I Li,dq ω0 is the angular frequency of the AC voltage and current in Park transformation, k i is the grid-connected coupling ratio, C ti and L ti is the equivalent capacitance and equivalent inductance at the grid connection point.

[0063] For two connected agents, the model of the transmission line is:

[0064] Line ij:

[0065] In the formula, the two agents are ij. ij,dq , L ij and R ij are the current, inductance and resistance parameters of the transmission line.

[0066] If approximation This is equivalent to excluding the disturbance caused by the current changes of other interconnected distributed generators in the network, and establishing the dynamic model of the generator node as follows:

[0067] DGU i:

[0068] For a single agent, the output information x i =[V i,d ,V i,q ] T , scheduling information u i =[V ti,d ,V ti,q ,I ti,d ,I ti,q ] T , the disturbance is ω i =[I Li,d ,I Li,q ] T The dispatch current is the real-time output current of the agent. By normalizing the output matrices of multiple agents, the distributed generator set can be dynamically modeled as a multi-agent system model containing N nodes.

[0069] Figure 2 Schematic diagram of a demand-driven distributed collaborative scheduling method in an unreliable network and a time-varying system of a single intelligent agent in the system according to the present invention. Figure 2 As shown, a multi-agent model consisting of multiple independent generator sets is constructed, including: the multi-agent model is:

[0070]

[0071] In the formula, x i (t) is the true output of the ith agent,

[0072] u i (t) is the scheduling input of the ith agent,

[0073] ω i (t) is the perturbation of the ith agent,

[0074] A, B and E are coefficient matrices, respectively. The coefficient matrices A, B and E are obtained in the process of normalizing the unit matrix.

[0075] The dispatch input of the ith agent is designed using the difference between the predicted states of the node itself and the neighboring nodes. The operation state of the distributed small power generation unit nodes of the local power plant is predicted, and we can get

[0076]

[0077] In the formula, is the predicted output of the jth generator node, is the set of all nodes adjacent to the i-th node. C is the state transition matrix.

[0078] Then the scheduling input of the i-th agent is expressed as:

[0079]

[0080] K is the scheduling gain matrix, and k is the number of the scheduling action. It can be seen that for each distributed generator set, there is a scheduling controller that uses the difference between the node's own predicted state and the neighboring node's predicted state to design a scheduling based on state feedback.

[0081] Solve the scheduling parameters of the consensus error system through the collaborative error convergence of multiple agents, including: Calculate the collaborative state when multiple agents are coupled And analyze the coordination error ε corresponding to the coordination state i (t); Utilize the collaborative error of each intelligent agent, the adjacency error between multiple intelligent agents, and the collaborative disturbance of each intelligent agent to construct a consistent error system for the collaborative error; Select the Lyapunov function of the consistent error system, and solve the scheduling parameters involved in the consistent error system according to the attenuation rate of the Lyapunov function; The scheduling parameters include the scheduling gain matrix K, the convergence coefficient matrix, the interference disturbance coefficient matrix, the first weight, and the second weight.

[0082] Define the average state of the generator network at the common coupling point:

[0083]

[0084] Where N is the total number of nodes, and the average of the actual output of each node is the average state of the network of the generator set. N is the unit matrix, then is the unit diagonal matrix.

[0085] The coordination error between any node i and the average state of the network is:

[0086]

[0087] The coordinated error of all nodes is the average of the errors of individual nodes, that is:

[0088]

[0089] in,

[0090] If the error of each node is Then the expression of the coordinated error matrix can also be solved as

[0091] Therefore, the dynamic system of collaborative scheduling error is established as:

[0092]

[0093] In the formula, A, B and E are still determined by the coefficient matrix of the multi-agent model, and the first term is the collaborative error, the second term is the influence from the adjacent nodes, and the third term is the error caused by the collaborative disturbance of each agent. K is obtained according to the scheduling gain matrix.

[0094] Γ(t) is the set of transmission lines within the attack interval, Indicates that all transmission lines used to calculate the coordination error are not in the attack interval, that is, this error is the expected amount of the system in the non-attack state. Specifically, the set of transmission lines that are in the attack interval and meet the demand-driven mechanism at any time t is defined as:

[0095]

[0096] Where ε is the set of all transmission lines in the system, Indicates the set of transmission attempts on channel (i, j) that are in the attack interval. Generally speaking, the elements in this set are discrete moments.

[0097] Then the cooperative error system in matrix form is:

[0098]

[0099] Where e(t) is the prediction error, and its elements are is the Laplace transform of the agent’s connectivity topology, is the Laplace transform of the topological matrix of the failed transmission line in the attack interval, if but The corresponding elements in are replaced with 0.

[0100] According to the Lyapunov stability analysis, a set of expressions about the convergence rate of the Lyapunov function of the set ε can be obtained, and the selection range of the relevant scheduling control parameters can be determined according to the convergence rate. In the present invention, a dynamic auxiliary variable is added to the Lyapunov function to dynamically adjust the Lyapunov function when the connectivity rate is not 0.

[0101]

[0102] In the formula, γ i (t) is a dynamic auxiliary variable, and through transformation, a set of Lyapunov function attenuation rates related to different transmission modes can be obtained. P is solved by (A+BK+E) T P+P(A+BK+E)=-Q is obtained, where Q is an arbitrary symmetry moment.

[0103] For any t∈Ψ Γ(t) , taking the first-order derivative of the Lyapunov function yields:

[0104]

[0105] Among them, Γ(t) is the time interval occupied by the transmission mode Γ(t), that is, the set of all scheduled moments. Γ is the Lyapunov decay rate, ξ(t)=[ε T (t)e T (t)] T ,

[0106] And there are:

[0107]

[0108] In addition, in order for the function to converge, it should be proved that for any given α i >0,σ i >0 and χ i (0)>0, dynamic auxiliary variable Υ i (t)>0. Among them, χ i (0) is the decision function, which will be given in the subsequent steps.

[0109] As mentioned above, is the Laplace matrix describing the topology of the generator network, for When , the elements are replaced by 0 Laplacian matrix. and Λ=diag{λ1,…,λ N}as well as m=[m2,…,m N ], through matrix transformation we can get:

[0110]

[0111] in,

[0112]

[0113] Using the linear matrix algorithm to solve, we get the positive definite symmetric matrix X that satisfies the above linear matrix inequality. and matrix Y, and through and Get the parameters to be solved.

[0114] In the derivation of linear matrix inequalities, it is necessary to prove that the designed demand-driven mechanism will not appear In the case of unlimited demand triggering, the calculation formula is required:

[0115]

[0116] According to the above characteristics, define π k , is the time point at which the set Γ(t) changes, then at t∈[π k ,π k+1 ) can be obtained:

[0117]

[0118] At the initial moment, when t>π0=0, we have:

[0119]

[0120] in, η * At least greater than or equal to 0.

[0121] Solving the scheduling parameters involved in the consistency error system according to the attenuation rate of the Lyapunov function includes: according to the preset equivalent attenuation rate φ of different transmission lines in the attack interval ij , the equivalent attenuation rate of different preset transmission lines in the non-attack interval Limit the decay rate of the Lyapunov function to β Γ The value range of .

[0122] Define the scalar φ ij and Any communication channel in the attack interval and non-attack interval The equivalent attenuation rate of , for a longer measurement period, is:

[0123]

[0124] in,

[0125] The equivalent attenuation rate φ of the different transmission lines in the attack interval is preset by the allowable attack intensity index ij , Equivalent attenuation rate of different transmission lines in the non-attack interval The permissible attack strength index The duration of the multi-agent model being attacked within a preset time interval with historical collection The correlation parameter τ ij related.

[0126] The unreliable network between nodes of distributed generators is modeled as a scheduling interruption caused by a malicious denial of service attack, and the attacks between any nodes are independent and time-varying, that is, the attacker can select any one or more transmission channels to carry out blocking attacks in any time interval. Assuming that the duration of the attack on any channel in any time interval is bounded, there exists a scalar κ ij ≥0 and τ ij >1Satisfy:

[0127]

[0128] in, For Channel The set of attack intervals in the time interval [t1, t2). The number of elements in this time set is related to the length of the time interval. If the two scalars limit the number of elements, it means that the attack is bounded, so the scheduling method based on the attack and the system error converge.

[0129] Depend on Available, when hour, and When:

[0130]

[0131] And the traditional Lyapunov function has more:

[0132]

[0133] Then we have:

[0134]

[0135] In the formula, ensure that ε(t) converges, and is the set of ε(t) values, and λ min (·) is the minimum value among the matrix eigenvalues, λ max (·) is the maximum value among the eigenvalues ​​of the matrix.

[0136] It can be seen that the above conditions and The bounded convergence of the collaborative scheduling error is guaranteed.

[0137] The equivalent attenuation rates of the different transmission lines in the attack interval and the equivalent attenuation rates of the different transmission lines in the non-attack interval are preset through the tolerable attack intensity index; the tolerable attack intensity index is related to the associated parameters of the duration of the attack on the multi-agent model collected historically within a preset time interval.

[0138] In one embodiment, φ ij and The permissible attack intensity index for the convergence of the consistency error system can be calculated based on the permissible recovery time of the transmission line and other parameters in advance, and the convergence characteristics of the received historical attacks. According to the value range of this index, the preset parameter φ can be adjusted. ij and The parameter φ can also be optimized according to the convergence effect of the results of the collaborative scheduling in the subsequent steps. ij and Such as improving the redundant backup capability, response speed and fault recovery speed of transmission lines.

[0139] Step 2: Use decision functions to describe the data update requirements to obtain the data interaction driving mechanism among multiple agents.

[0140] The decision function is used to describe the data update requirements to obtain a data interaction driving mechanism between multiple agents including scheduling parameters, including: the decision function is:

[0141]

[0142] The data interaction driving mechanism between the multiple agents is:

[0143]

[0144] In the formula, γ i (t) is the dynamic auxiliary variable of the ith agent at time t,

[0145] Ω iis the convergence coefficient of the ith agent,

[0146] θ i is the interference interaction coefficient of the ith agent, and

[0147] and are the predicted outputs between the ith agent and the jth agent at time t,

[0148] a ij is the connectivity coefficient between the ith agent and the jth agent,

[0149] N is the total number of agents,

[0150] is the time when the i-th agent requests scheduling for the kth time, and

[0151] The difference between the actual state and the predicted state of each node is used as a quantitative indicator of demand, and a demand-driven transmission mechanism for data interaction between scheduling nodes is constructed.

[0152] The decision function describes the data update requirements in the prediction function, and the convergence coefficient matrix Ω i and the phase interference coefficient matrix θ i Given by stability analysis, the phase interference coefficient matrix can be related to the gain coefficient matrix.

[0153] The dynamic auxiliary variable system is:

[0154]

[0155] in,

[0156] σ i >0 is the second weight, α i is the first weight, a ij are the elements in the connectivity matrix.

[0157] Step 3: Determine whether there is any transmission attempt within the attack interval, and implement distributed collaborative scheduling of multiple agents for the attack transmission attempt based on the data interaction driving mechanism and the secure transmission scheduling protocol.

[0158] Determine whether there is a transmission attempt in the attack interval, and implement distributed collaborative scheduling of multiple agents for the transmission attempt according to the data interaction driving mechanism and the secure transmission scheduling protocol, including: the time series of multiple attack interval transmission attempts between the i-th agent and the j-th agent is:

[0159]

[0160] In the formula, The set of allowed time for request scheduling of the i-th agent;

[0161] like Then respond to the current scheduling request and adjust the scheduling input of the i-th agent based on the controller gain; if Then delay the current scheduling request to Δ i is the preset time interval.

[0162] After determining the attack attempt, a secure transmission scheduling protocol can be designed as:

[0163]

[0164] In the formula, The actual meaning is to execute scheduling according to the current scheduling instruction. The preset time interval is the minimum transmission interval or an integer multiple of the minimum transmission interval.

[0165] On this basis, the scheduling controller is obtained as:

[0166]

[0167] Figure 3 The present invention is a schematic diagram of a multi-agent connection mode in a demand-driven distributed collaborative scheduling method and system in an unreliable network. In this embodiment, there are three transmission lines between the grid connection points of four distributed generator sets.

[0168] Figure 4 The diagram is a schematic diagram of independent denial of service attacks on different connected edges in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. When there is no distributed collaborative scheduling, each connected edge needs to respond excessively for a denial of service attack.

[0169] Fig. 9 Schematic diagram of input control signals of four agents in an embodiment of a distributed collaborative scheduling method and system based on demand-driven in an unreliable network of the present invention. Fig. 9 As shown in the figure, the scheduling control is input to the four agents respectively. The scheduling control received by the four agents is different.

[0170] Figure 5 It is a schematic diagram of a system operation trajectory monitored by a first intelligent agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. Figure 6 It is a schematic diagram of a system operation trajectory monitored by a second intelligent agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. Figure 7 It is a schematic diagram of a system operation trajectory monitored by a third intelligent agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. Figure 8 It is a schematic diagram of a system operation trajectory monitored by a fourth intelligent agent in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention.

[0171] like Figure 5-8 ,Through coordinated response, the over-limit problem on a certain transmission line was reduced, the ,devices compensated for each other, and the system was stable within the ,preset attenuation rate range.

[0172] Fig.10 The discrete distribution diagram of the scheduling requests of four intelligent agents in an embodiment of a demand-driven distributed collaborative scheduling method and system in an unreliable network of the present invention. Fig.10 As shown in Figure 1, each agent has different degrees of scheduling requests at different decentralized moments within 10 seconds.

[0173] Fig.11 This is an evolution diagram of dynamic auxiliary variables of four agents in an embodiment of a distributed collaborative scheduling method and system based on demand-driven in an unreliable network of the present invention. Within 10 seconds, the state of the multi-agent model gradually converges.

[0174] The second aspect of the present invention relates to a demand-driven distributed collaborative scheduling system in an unreliable network; the system is implemented using a demand-driven distributed collaborative scheduling method in an unreliable network in the first aspect of the present invention; wherein the system includes a building module, a description module and a scheduling module; wherein the building module is used to construct a multi-agent model composed of multiple independent individual generator sets, and solve the multi-agent model through the collaborative error convergence of the multi-agent; the description module is used to use a decision function to describe the demand for data update to obtain a data interaction driving mechanism between the multi-agents; the scheduling module is used to determine whether there is a transmission attempt within the attack interval, and implement distributed collaborative scheduling of the multi-agents for the transmission attempt according to the data interaction driving mechanism and the secure transmission scheduling protocol.

[0175] A third aspect of the present invention relates to a terminal, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method described in the first aspect of the present invention.

[0176] A fourth aspect of the present invention relates to a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the method in the first aspect of the present invention.

[0177] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention is described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the technical solutions of the present invention still include contents that can modify or replace the specific embodiments of the present invention. Any modification or replacement that does not depart from the spirit and scope of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A demand-driven distributed collaborative scheduling method in an unreliable network, characterized in that: The method comprises the following steps: Construct a multi-agent model consisting of multiple independent generators, and solve the scheduling parameters of the consensus error system through the collaborative error convergence of the multi-agents; The decision function is used to describe the data update requirements to obtain the data interaction driving mechanism between multiple agents including scheduling parameters; Determine whether there is a transmission attempt within the attack interval, and implement distributed collaborative scheduling of multiple agents for the attack transmission attempt based on the data interaction driven mechanism and the secure transmission scheduling protocol.

2. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 1, characterized in that: The multi-agent model consisting of multiple independent generator sets is constructed, including: The multi-agent model is: In the formula, x i (t) is the true output of the ith agent, u i (t) is the scheduling input of the ith agent, ω i (t) is the perturbation of the ith agent, A, B and E are coefficient matrices respectively.

3. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 2, characterized in that: The method of solving the scheduling parameters of the consistency error system through the collaborative error convergence of multiple agents includes: Calculating the collaborative state of multi-agent coupling And analyze the coordination error ε corresponding to the coordination state i (t); Using the collaborative error of each intelligent agent, the adjacency error between multiple intelligent agents, and the collaborative disturbance of each intelligent agent, a consistent error system for the collaborative error is constructed; Selecting a Lyapunov function of the consistency error system, and solving a scheduling parameter involved in the consistency error system according to a decay rate of the Lyapunov function; The scheduling parameters include a scheduling gain matrix, a connectivity matrix, a convergence coefficient matrix, a phase interference coefficient matrix, a first weight, and a second weight.

4. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 3, characterized in that: Solving the scheduling parameters involved in the consistency error system according to the attenuation rate of the Lyapunov function includes: The value range of the attenuation rate of the Lyapunov function is limited according to the preset equivalent attenuation rates of different transmission lines in the attack interval and the preset equivalent attenuation rates of different transmission lines in the non-attack interval.

5. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 4, characterized in that: The equivalent attenuation rates of the different transmission lines in the attack interval and the equivalent attenuation rates of the different transmission lines in the non-attack interval are preset by the allowable attack intensity index; The tolerable attack intensity index is related to the associated parameters of the duration of the multi-agent model being attacked within a preset time interval collected historically.

6. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 5, characterized in that: The use of decision functions to describe the need for data update to obtain a data interaction driving mechanism between multiple agents including scheduling parameters includes: The decision function is: The data interaction driving mechanism between the multiple agents is: In the formula, γ i (t) is the dynamic auxiliary variable of the ith agent at time t, e i (t) is the prediction error, that is, Ω i is the convergence coefficient of the ith agent, θ i is the interference interaction coefficient of the ith agent, and and are the predicted outputs between the ith agent and the jth agent at time t, a ij is the connectivity coefficient between the ith agent and the jth agent, N is the total number of agents, is the time when the i-th agent requests scheduling for the kth time.

7. The demand-driven distributed collaborative scheduling method in an unreliable network according to claim 6, characterized in that: The determining whether there is a transmission attempt within the attack interval, and implementing distributed collaborative scheduling of multiple agents for the transmission attempt according to the data interaction driving mechanism and the secure transmission scheduling protocol, includes: The time series of multiple attack interval transmission attempts between the i-th agent and the j-th agent is: In the formula, The set of allowed time for request scheduling of the i-th agent; like Then respond to the current scheduling request and adjust the scheduling input of the i-th agent based on the controller gain; like Then delay the current scheduling request to is the preset time interval.

8. A demand-driven distributed collaborative scheduling system in an unreliable network; characterized in that: The system is implemented using a demand-driven distributed collaborative scheduling method in an unreliable network as described in any one of claims 1 to 7; wherein, The system includes a construction module, a description module and a scheduling module; wherein, The construction module is used to construct a multi-agent model composed of a plurality of independent generator sets, and solve the multi-agent model through the collaborative error convergence of the multi-agents; The description module is used to describe the demand for data update by using a decision function to obtain a data interaction driving mechanism between multiple agents; The scheduling module is used to determine whether there is a transmission attempt within the attack interval, and implement distributed collaborative scheduling of multiple agents for the transmission attempt based on the data interaction driving mechanism and the secure transmission scheduling protocol.

9. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images