Direct-current power distribution network information physical cooperation active power control method and device oriented to multi-microgrid access

Abstract

The invention discloses a design method of a DC power distribution network information physical cooperation active power control method and device oriented to multi-microgrid access. The design method comprises the following steps: formulating an information physical cooperation correlation analysis and cooperation active power regulation and control framework; designing a centralized control method of a physical layer according to power flow constraint on an upper layer, namely a micro-grid layer; designing a combined fault-tolerant control method at a lower layer, namely a distributed resource layer; the effectiveness of the method is verified by building an experimental scene. According to the method, under the condition that the direct-current power distribution network under multi-micro-grid access can cause multiple uncertainty of information physics, fluctuation of active power can be effectively restrained, data transmission path matching based on service requirements is achieved, and the method has higher active anti-interference capacity.

Description

Technical Field

[0001] The present invention relates to the field of smart grid control, and particularly relates to a method and device for cyber-physical collaborative active power control of a DC distribution network under the access of multiple microgrids. Background Art

[0002] As one of the core components of the smart grid, the DC distribution network has advantages such as strong regulation flexibility and high efficiency, attracting more and more attention. In the DC distribution network, active power regulation is one of its key technologies, which is of great significance for ensuring the safe and stable operation of the power grid, improving the operation efficiency of the power grid, and optimizing energy utilization. In recent years, with the increasing proportion of aggregated distributed resources such as microgrids connected to the whole station acceleration network, the research on the flexible deployment of microgrid clusters and their internal controllable resources has become a research hotspot.

[0003] In the traditional AC distribution network, active power regulation is achieved by adjusting the output power of generators and the current of loads. However, in the DC distribution network, due to the uniqueness of the current direction and the fixity of the voltage on the DC line, active power can only be achieved by adjusting the voltage on the DC side, which leads to a lack of flexibility in the active power regulation mechanism. However, in recent years, with the increasing proportion of distributed resource aggregates such as microgrids accessing the distribution network, researching the regulation potential of flexibly invoking microgrids and their internal controllable resources has become a research hotspot. Moreover, considering the disordered disturbance characteristics of a large number of distributed resources themselves, data acquisition and transmission links are often added to the distribution network to achieve real-time monitoring and analysis of key information such as operating status and load demand, improving the overall observability and controllability of the distribution network and microgrids. This also makes the distribution network gradually show the characteristics of cyber-physical integration.

[0004] However, it cannot be ignored that in the cyber-physical integration environment, physical layer uncertainties such as source-side output fluctuations and information layer uncertainties such as communication congestion will pose severe challenges to the active power regulation of the DC distribution network. On the one hand, due to the relatively significant fluctuations in the output of distributed power sources accessed by microgrids in the DC distribution network, for example, the volatility of renewable energy is relatively large, and situations such as sudden shutdowns or increased outputs may occur from time to time. Such fluctuations will directly affect the active power regulation of the DC distribution network. In this case, traditional static power system scheduling methods may no longer be applicable, and more flexible and faster-response scheduling strategies are needed to cope with the challenges brought by output fluctuations. On the other hand, communication congestion is another important challenge faced by the DC distribution network. In the information transmission link, due to a large amount of data transmission and communication leading to network congestion, it may cause delayed transmission or loss of active power regulation information, thereby affecting the real-time monitoring and control of the DC distribution network.

[0005] Therefore, the present invention proposes a cyber-physical collaborative active power control method and device for a DC distribution network with multiple microgrid accesses, which can maintain the active power balance of the distribution network under multiple cyber-physical uncertainties. Summary of the Invention

[0006] To solve the above problems, the present invention discloses a cyber-physical collaborative correlation analysis and coordinated control device that can maintain the active power balance of the distribution network from both cyber and physical perspectives in the case of active power fluctuations caused by multiple new energy sources accessing the distribution network.

[0007] To achieve the above objectives, the following technical solutions are adopted: for the upper layer, i.e., the microgrid layer, a method for supporting the active power of the microgrid is proposed in the physical layer, and an optimal path matching method is proposed in the network layer; for the lower layer, i.e., the distributed resource layer, a multi-resource distributed fault-tolerant active power control method is proposed in the physical layer, and an event-triggered active defense method is proposed in the network layer.

[0008] The specific steps are as follows:

[0009] Step 1, formulate a cyber-physical collaborative correlation analysis and coordinated active power regulation architecture.

[0010] Step 2, design a centralized control method based on power flow constraints in the upper layer.

[0011] Step 3, design a combined fault-tolerant control method in the lower layer.

[0012] Step 4, verify the effectiveness of the method by building an experimental scenario.

[0013] Furthermore, in Step 1, the main components of the DC distribution network can be divided into two layers: the physical space and the information space. The overall structure design is proposed from the idea of "cyber-physical collaborative correlation analysis and cooperation", where "cyber-physical collaboration" includes the network layer and the physical layer; "correlation analysis" means that the physical layer knows the network situation and the network layer knows the requirements of physical services; "cooperation" means that the network layer should provide reliable communication support according to the service requirements of the physical layer; and the control method of the physical layer should be adaptively adjusted according to the support capabilities that the network layer can provide.

[0014] Furthermore, in Step 2, after completing the design of the overall framework diagram, it is the design of the matching strategy for the wired network architecture of the information layer and the aggregated regulation strategy of the physical layer based on the cyber-physical correlation analysis and coordination idea, that is, the upper layer design. The design is carried out from the perspectives of cyber-physical correlation analysis and coordination respectively. The following is the design process.

[0015] Step 2-1, cyber-physical correlation analysis

[0016] 1) Analyze the physical layer service requirements in the information layer

[0017] Model the correlation between the execution effect of active power regulation and the transmitted data; further, based on the sensitivity analysis method, quantify the impact on the active power regulation effect under the changes of data transmission speed and accuracy to determine the importance of different data. When facing service G, when its relationship with relevant parameters can be modeled as

[0018] G = f(x 1 ,…,x m ) (1)

[0019] where x 1 ,…,x m is the data set and f is the mapping relationship; then the impact of the change in the transmission speed and accuracy of the i-th parameter on the execution effect of service G can be expressed by sensitivity, overshoot, and steady-state error.

[0020] 2) Physical layer correlation analysis The network environment provided by the information layer

[0021] After the network matching is completed, information about the router status, network architecture, and transmission channel status will be fed back to the physical layer. Based on the above data, the physical layer can adjust the active power control strategy in real time. In this way, by adjusting the control strategy accordingly, the system can more flexibly adapt to different environments..

[0022] Step 2-2, Information-physical collaborative coordination

[0023] 1) Power flow constraint-dependent centralized control

[0024] First, the supply-demand relationship and power flow are analyzed; in addition, for the source-network energy storage microgrid, multiple factors are comprehensively considered to formulate control instructions with the goal of the shortest regulation and load balancing.

[0025] 1-1) Optimization model construction

[0026] Given that the power fluctuation amount is m microgrids, for the i-th microgrid, the control command is set as a multi-objective (maximizing load balancing, quickly suppressing power fluctuations, minimizing line losses, minimizing voltage deviation, etc.) optimization model as

[0027]

[0028] where △P α / P α and △P β / P β are the ratios of the active power changes of the α-th and β-th microgrids to their capacities; Speed 1 and Speed m are the response speeds; △P α , △Pβ and △P m is the active power change; P loss.α is the line loss of the α-th microgrid; △U α is the voltage fluctuation between the α-th buses; || abs represents taking the absolute value; || Nor represents the normalization operation; ρ 1 +ρ 2 +ρ 3 +ρ 4 = 1. The relevant constraints are power balance: and adjustable capacity limit: 0 ≤ △P α ≤ △P α.max .

[0029] 1 - 2) Constraint Conditions

[0030] In the present invention, some objectives are selected to be transformed into constraints. For example,

[0031] Response time constraint: max(|△P 1 / Speed 1 | Nor , …, |△P m / Speed m | Nor ) ≤ T max (3)

[0032] Line loss constraint:

[0033] Node voltage limit: U α.min ≤ U α ≤ U α.max (5)

[0034] where △P α is the output active power of the α-th microgrid; △P α.max is the maximum adjustable power of the α-th microgrid; △P α.loss is the line loss of the α-th microgrid; P lineloss.max is the set loss threshold; U α.min and U α.max are the minimum and maximum voltages of the node; T max is the set maximum response time.

[0035] 2) Demand-Driven Wired Network Matching

[0036] 2 - 1) Optimization Model Construction

[0037] First, the sensitivity calculation should be completed in the active power control. Assuming there are m buses, the sensor static pose equation for the power injection change in the α-th bus in objective (2) is

[0038]

[0039]

[0040] where is the partial derivative of the output power of Z with respect to the α-th microgrid in (2). After determining the sensitivity, the flow data to be transmitted is sorted in descending or ascending order of sensitivity to obtain

[0041]

[0042] In addition, the centralized tree network can be planned according to the order in (7). The related objective function is constructed as

[0043]

[0044] where Sign is the sign function; Per αβ is the performance of the channel from the α-th router (connected to α microgrids) to the β-th router (connected to β microgrids) (e.g., delay, length, security risk, etc.); a αβ is the relationship between the α-th and β-th routers. If the α-th router can send data to the β-th router, a αβ = 1 or a αβ = 0; otherwise, N α is the set of neighbor routers of the α-th router.

[0045] 2-2) Constraint conditions

[0046] The constraints of the upload network are constructed as follows:

[0047] For the α-th router, the number of input channels and output channels should not exceed the allowed values:

[0048]

[0049] The number of connected channels >= the number of microgrids:

[0050]

[0051] Each transmission router has its output channels and input channels:

[0052]

[0053] Each microgrid has at least one output channel:

[0054]

[0055] The microgrid control center should be connected to at least the same number of input channels as the microgrid:

[0056]

[0057] For the download network, constraints (12) and (13) are modified to the following equations, while other constraints remain unchanged:

[0058] Each MG has at least one input channel:

[0059]

[0060] The center has at least one down-load channel:

[0061]

[0062] where Num out.α and Num in.α are the numbers of output and input channels allowed for the α-th node; N Tr is the set of transmission routers. Once the above optimization problem is resolved again, the paths can be obtained in sequence.

[0063] Furthermore, in step 3, after the upper layer design is completed, the design of the active power regulation strategy with active fault tolerance and the dual-path optimization strategy for the lower layer is carried out.

[0064] 3-1) Design of combined fault-tolerant control method

[0065] Based on the control command of the microgrid, the present invention can adaptively adjust the controller parameters based on the equal capacity ratio P h-α / P h-α.cap =P h-β / P h-β.cap (representing the ratio of the output command to the rated capacity of the α / β distributed resources) and the control command of "virtual leader-follower consensus control", so that the resources can flexibly respond to the commands of the microgrid. The present invention takes an example to illustrate that the control equation for the performance of the α-th resource is

[0066] △P h-α =△P h-α.Ref (16)

[0067] where △P h-α and △P h-β are the active power adjustments of the α-th and β-th distributed resources in the h-th microgrid; P h-α.cap and P h-β.cap are the remaining capacities of the α-th and β-th resources; △P h-α.Ref and △P h-β.Ref are obtained through the inverse solution of the control result (i.e., △P h-α.Ref =γ h-α (t)·Ph-α.cap and △P h-β.Ref = γ h-β (t)·P h-β.cap ) is obtained. In this case, γ h-L (t) is usually selected as where P MGh is the power required by the h-th microgrid. The relevant performance equation of the distributed resource is

[0068]

[0069] The output command is

[0070]

[0071] where u h-1 (t) and u h-n (t) are controllers; k h-α1 , k h-α2 , k h-n2 and k h-n1 are gains; a h-αβ represents the correlation between the α-th and β-th resources; b h-α represents the association between the α-th resource and the volume load. During the regulation process, the deviation between the controlled variable and the set value of the volume load is e h = u h - γ h-L . Under normal circumstances, the controlled deviation is adjusted to 0 by selecting appropriate parameters. However, if the network switches, the resulting interference will have a negative impact on the performance quality, and sliding mode control is required to suppress it.

[0072] Taking a system as an example, the equations of the affected controllers u h ′ -1 (t) ~ u h ′ -n (t) are (19), where △γ h-1 (t) and △γ h-n (t) are the introduced interferences.

[0073]

[0074] The controllable error u′ -u is

[0075]

[0076] where △u h-1 (t) = u h ′ -1 (t) - u h-1 (t). In the present invention, the feedback matrix F is used to complete the sliding mode control, and it is designed as where F h-αβ = -K γh-αβ a h-αβ ; F h-αα needs to be designed. Taking the α-th distributed resource as an example, the sliding surface is designed as S = K P △γ h-α + K I ∫△γ h-α dt, where and are coefficients. The relevant controller is designed as F h-αα = F h-αa + F h-αb . Let and F h-α1 = F h-αa , we can get

[0077]

[0078] where F h-αa is the solution of equation (21). Then, taking the Lyapunov function as L = S 2 / 2 and making it satisfy Lyapunov stability, considering F h-αα = F h-αa + F h-αb , we can get where is the switching gain in the protocol. Therefore, when selecting appropriate parameters, the tolerance of the system to congestion can be improved.

[0079] 3 - 2) Design of the dual - path optimization strategy

[0080] For the distributed resources in the micro - grid, in order to support the execution of distributed control, it is necessary to ensure the existence of a directed spanning tree in the network. To this end, relying on the micro - grid agent and router, global optimization in the warning state and local optimization in the emergency state will be carried out respectively to form a dual - optimization strategy. The warning state refers to the situation where the inequality constraint is close to the limit, making it easy to transition to the emergency state; the emergency state refers to the situation where the inequality constraint has been violated and it is difficult to satisfy the equality constraint, and emergency power support is required to suppress fluctuations. The design process is as follows:

[0081] First, when the micro - grid is operating in the warning state, the sensitivity between the controlled variable and the relevant data should be calculated; when the system is operating in the emergency state, the sensitivity between the controlled variable and time should be calculated. Taking the α - th distributed resource as an example, the partial derivative between its controller and the data of the β - th distributed resource is Its partial derivative with respect to time is

[0082]

[0083] In addition, the path needs to be updated. The upper-layer optimization mainly solves the problems in the alert state. Based on the principle of matching path performance with data importance, a directed spanning tree is reconstructed using global information. When the α-th distributed resource needs to be connected to the directed tree, the transmission path should match the data in sequence. The optimization objective is The constraint conditions are:

[0084] The number of channels the number of DERs:

[0085]

[0086] There is at most one connected channel between two routers:

[0087]

[0088] The controlled DER has only one input channel:

[0089]

[0090] The volume load has at least one output channel and no input channel:

[0091]

[0092] Where P αβ is the security probability of the channel from the β-th router to the α-th router; Number DER is the number of distributed resources that need to be connected.

[0093] The lower-layer optimization mainly solves the path reconstruction problem in the emergency state. Based on the transmission network constructed by the upper layer, first, the selected paths are set to 1, and the unselectable paths and faulty paths are set to 0. In addition, based on the principle of matching transmission speed and data input, the distributed resources use local information to quickly generate a path reconstruction plan. The target model is Where t αβ is the transmission time of the channel from the α-th distributed resource to the β-th distributed resource. If the i-th distributed resource needs to be connected to the network. The following constraints need to be considered in the optimization model;

[0094] Transmission security constraint:

[0095] ∏P αβ ·Sign(a αβ )≥Prob set (27)

[0096] Ensure that the i-th distributed resource will reconnect to the directed tree:

[0097]

[0098] where Prob set is the required safety probability. If it is necessary to reconnect multiple sets of data, the data should be matched with the channels in sequence.

[0099] Furthermore, in step 4, the effectiveness of the method is verified by building an experimental scenario.

[0100] Compared with the prior art, the present invention has the following advantages:

[0101] 1. The centralized control method of the present invention generates the optimal control command based on the power flow constraint, comprehensively considering factors such as remaining capacity, regulation time, voltage constraint, and line loss. Compared with the traditional equal capacity ratio method, it is more flexible and applicable to various operating scenarios.

[0102] 2. The path matching algorithm can calculate the data sensitivity and match the optimal path, realizing the data transmission path matching in a service-driven manner, which is superior to the traditional method.

[0103] 3. The combined fault-tolerant control method can achieve active power regulation only relying on a sparse communication network. Compared with the common centralized control methods, the proposed method greatly reduces the communication cost and calculation burden.

[0104] 4. The dual optimization method can perform global and local optimizations to meet the requirements of multi-scenario services. BRIEF DESCRIPTION OF THE DRAWINGS

[0105] Figure 1 Architecture diagram of information-physical collaborative association analysis and coordinated active power regulation for DC distribution network.

[0106] Figure 2 Schematic diagram of centralized regulation for microgrid clusters.

[0107] Figure 3 Schematic diagram of multi-source distributed collaborative regulation. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0108] The present invention will be further clarified below in conjunction with the drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. It should be noted that the terms "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to the directions in the drawings, and the terms "inner" and "outer" respectively refer to the directions towards or away from the geometric center of a specific component.

[0109] As Figure 1As shown in the figure, this embodiment designs an information - physical collaborative association analysis and cooperative active power regulation architecture diagram for a DC distribution network, including an upper - layer wired network, a lower - layer wireless network, a transmission line, a micro - grid control center, a micro - grid group, and distributed resources. When there is an active power fluctuation in the DC distribution network, first, based on information - physical collaboration, the association mechanism is clarified. By designing a hierarchical control strategy and a reliable transmission scheme, the power control capabilities of the micro - grid and its internal distributed resources are utilized in an information - physical cooperation manner to eliminate the fluctuation.

[0110] As Figure 2 and 3 shown, the present invention provides an information - physical collaborative active power control method for a DC distribution network under multi - micro - grid access, including:

[0111] S1. Formulate an information - physical collaborative association analysis and cooperative active power regulation architecture

[0112] Design the information - physical collaborative association analysis and cooperative active power regulation architecture of the DC distribution network from the idea of "information - physical collaborative association analysis and cooperation", where "information - physical collaboration" includes a network layer and a physical layer; "association analysis" means that the physical layer knows the network situation and the network layer knows the requirements of physical services; "cooperation" means that the network layer should provide reliable communication support according to the service requirements of the physical layer; the control method of the physical layer should be adaptively adjusted according to the support capabilities that the network layer can provide.

[0113] S2. Design a centralized control method based on power flow constraints in the upper layer

[0114] As Figure 2 shown, after completing the information - physical collaborative association analysis and cooperative active power regulation architecture, the upper - layer design is carried out, and the specific steps are as follows:

[0115] First, the supply - demand relationship and power flow are analyzed; in addition, for the source - network - energy - storage micro - grid, various factors are comprehensively considered, and control instructions are formulated with the goal of the shortest regulation and load balancing.

[0116] 1 - 1) Optimization model construction

[0117] Given that the power fluctuation amount is for m micro - grids, for the i - th micro - grid, the control command is set as a multi - objective (maximizing load balancing, quickly suppressing power fluctuations, minimizing line losses, minimizing voltage deviation, etc.) optimization model as

[0118]

[0119] where △P α / P α and △P β / P βis the ratio of the active power change of the α-th and β-th microgrids to their capacities; Speed 1 and Speed m is the response speed; △P α , △P β and △P m is the active power change; P loss.α is the line loss of the α-th microgrid; △U α is the voltage fluctuation between the α-th buses; || abs represents taking the absolute value; || Nor represents the normalization operation; ρ 1 +ρ 2 +ρ 3 +ρ 4 =1. The relevant constraints are power balance: and adjustable capacity limit: 0 ≤ △P α ≤ △P α.max .

[0120] 1-2) Constraint conditions

[0121] In the present invention, some objectives are selected to be transformed into constraints. For example,

[0122] Response time constraint: max(|△P 1 / Speed 1 | Nor , …, |△P m / Speed m | Nor ) ≤ T max (3)

[0123] Line loss constraint:

[0124] Node voltage limit: U α.min ≤ U α ≤ U α.max (5)

[0125] where △P α is the output active power of the α-th microgrid; △P α.max is the maximum adjustable power of the α-th microgrid; △P α.loss is the line loss of the α-th microgrid; P lineloss.max is the set loss threshold; U α.min and U α.max are the minimum and maximum voltages of the node; T max is the set maximum response time.

[0126] 2) Demand-driven wired network matching

[0127] 2-1) Optimization model construction

[0128] First, the sensitivity calculation should be completed in the active power control. Assume there are m buses, and the sensor static pose equation for the power injection change in the α-th bus on the target (2) is

[0129]

[0130] where is the partial derivative of Z with respect to the output power of the α-th microgrid in (2). After determining the sensitivity, the flow data to be transmitted is sorted in descending or ascending order of sensitivity to obtain

[0131]

[0132] In addition, the centralized tree network can be planned according to the order in (7). The relevant objective function is constructed as

[0133]

[0134] where Sign is the sign function; Per αβ is the performance of the channel from the α-th router (connected to α microgrids) to the β-th router (connected to β microgrids) (e.g., delay, length, security risk, etc.); a αβ is the relationship between the α-th and β-th routers. If the α-th router can send data to the β-th router, a αβ = 1 or a αβ = 0; otherwise, N α is the set of neighbor routers of the α-th router.

[0135] 2-2) Constraint conditions

[0136] The constraints of the upload network are constructed as:

[0137] For the α-th router, the number of input channels and output channels should not exceed the allowed values:

[0138]

[0139] The number of connected channels >= the number of microgrids:

[0140]

[0141] Each transmission router has its output channels and input channels:

[0142]

[0143] Each microgrid has at least one output channel:

[0144]

[0145] The microgrid control center shall be connected to at least the same number of input channels as the microgrid:

[0146]

[0147] For the download network, constraints (12) and (13) are modified to the following equations, while other constraints remain unchanged:

[0148] Each MG shall have at least one input channel:

[0149]

[0150] The center shall have at least one down - load channel:

[0151]

[0152] where Num out.α and Num in.α are the numbers of output and input channels allowed for the α - th node; N Tr is the set of transmission routers. Once the above - mentioned optimization problem is resolved again, the paths can be obtained in sequence.

[0153] S3. Design a combined fault - tolerant control method at the lower layer

[0154] As Figure 3 shown, the lower - layer design adopts a combined fault - tolerant control method, and the specific steps are as follows:

[0155] According to the control command of the microgrid, based on the equal - capacity ratio P h-α / P h-α.cap =P h-β / P h-β.cap (representing the ratio of the output command to the rated capacity of the α / β distributed resources) and the control command of "virtual leader - follower consensus control", this invention can adaptively adjust the controller parameters to enable the resources to flexibly respond to the commands of the microgrid. In this embodiment, the control equation for the performance of the α - th resource is

[0156] △P h-α =△P h-α.Ref (16)

[0157] where △P h-α and △P h-β are the active - power adjustments of the α - th and β - th distributed resources in the h - th microgrid; P h-α.cap and P h-β.cap are the remaining capacities of the α - th and β - th resources; △P h-α.Ref and △P h-β.Ref are obtained through the inverse solution of the control result (i.e., △Ph-α.Ref = γ h-α (t)·P h-α.cap and △P h-β.Ref = γ h-β (t)·P h-β.cap ) obtained. In this case, γ h-L (t) is usually selected as where P MGh is the power required by the h-th microgrid. The relevant performance equation of the distributed resource is

[0158]

[0159] The output command is

[0160]

[0161] where u h-1 (t) and u h-n (t) are controllers; k h-α1 , k h-α2 , k h-n2 and k h-n1 are gains; a h-αβ represents the correlation between the α-th and β-th resources; b h-α represents the association between the α-th resource and the volumetric load. During the regulation process, the deviation between the controlled variable and the set value of the volumetric load is e h = u h - γ h-L . Under normal circumstances, the controlled deviation is adjusted to 0 by selecting appropriate parameters. However, if the network switches, the resulting interference will have a negative impact on the performance quality, and sliding mode control is required to suppress it. Taking a system as an example, the equations of the affected controllers u′ h-1 (t) ~ u′ h-n (t) are (19), where △γ h-1 (t) and △γ h-n (t) are the introduced interferences.

[0162]

[0163] The controllable error u′ - u is

[0164]

[0165] where △u h-1 (t) = u h ′ -1 (t) - u h-1 (t). In the present invention, the feedback matrix F is used to complete the sliding mode control, and it is designed as where F h-αβ = -Kγh-αβ a h-αβ ; F h-αα needs to be designed. Taking the α-th distributed resource as an example, the sliding surface is designed as S = K P △γ h-α + K I ∫△γ h-α dt, where and are coefficients. The related controller is designed as F h-αα = F h-αa + F h-αb . Let and F h-α1 = F h-αa , we can get

[0166]

[0167] where F h-αa is the solution of equation (21). Then, the Lyapunov function is taken as L = S 2 / 2, and make it satisfy Lyapunov stability. Considering F h-αα = F h-αa + F h-αb , we can get where is the switching gain in the protocol. Therefore, when selecting appropriate parameters, the tolerance of the system to congestion can be improved.

[0168] The design process of the dual optimization path is as follows:

[0169] First, when the microgrid operates in the alarm state, the sensitivity between the controlled variable and the relevant data should be calculated; when the system operates in the emergency state, the sensitivity between the controlled variable and time should be calculated. Taking the α-th distributed resource as an example, the partial derivative between its controller and the data of the β-th distributed resource is Its partial derivative with respect to time is

[0170]

[0171] In addition, the path needs to be updated. The upper-layer optimization mainly solves the problems in the alarm state. Based on the principle of matching path performance and data importance, the global information is used to reconstruct the directed spanning tree. When the α-th distributed resource needs to be connected to the directed tree, the transmission path should be matched with the data in sequence. The optimization objective adopts The constraint condition is:

[0172] The number of channels The number of DERs:

[0173]

[0174] There is at most one connected channel between two routers:

[0175]

[0176] The controlled DER has only one input channel:

[0177]

[0178] The volumetric load has at least one output channel and no input channels:

[0179]

[0180] where P αβ is the security probability of the channel from the β-th router to the α-th router; Number DER is the number of distributed resources to be connected.

[0181] The lower-layer optimization mainly solves the path reconstruction problem in emergency situations. Based on the transmission network constructed by the upper layer, it first sets the selected path to 1 and the non-selectable paths and faulty paths to 0. In addition, based on the principle of matching the transmission speed and data input, the distributed resources use local information to quickly generate a path reconstruction scheme. The objective model is where t αβ is the transmission time of the channel from the α-th distributed resource to the β-th distributed resource. If the i-th distributed resource needs to be connected to the network. The following constraints need to be considered in the optimization model;

[0182] Transmission security constraint:

[0183] ∏P αβ ·Sign(a αβ )≥Prob set (27)

[0184] Ensure that the i-th distributed resource will reconnect to the directed tree:

[0185]

[0186] where Prob set is the required security probability. If multiple groups of data need to be reconnected, the data should be matched with the channels in sequence.

[0187] S4. Verify the effectiveness of this method by building an experimental scenario.

[0188] Table 1 Microgrid Inter-channel Communication Performance

[0189]

[0190] Example 2

[0191] Based on the active power control method for cyber-physical cooperation in a DC distribution network under the access of multiple microgrids described in Embodiment 1, an embodiment of the present invention provides an active power control device for cyber-physical cooperation in a DC distribution network under the access of multiple microgrids, including:

[0192] An overall structure module, configured to: formulate an architecture for cyber-physical cooperation association analysis and collaborative active power regulation.

[0193] An upper-layer design module, configured to: design a centralized control method based on power flow constraints.

[0194] A lower-layer design module, configured to: design a combined fault-tolerant control method.

[0195] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0196] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in Figure 1 one or more flows and / or Figure 1 blocks or multiple blocks.

[0197] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes the functions specified in Figure 1 one or more flows and / or Figure 1 blocks or multiple blocks.

[0198] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to generate a computer-implemented process, thereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 steps for implementing the functions specified in one block or multiple blocks.

[0199] The technical means disclosed by the solution of the present invention are not limited to the technical means disclosed in the above embodiments, and also include technical solutions composed of any combination of the above technical features.

Claims

1. A method for controlling active power of a DC distribution network with cyber-physical collaboration under multi-microgrid access, characterized in that: The specific steps include: Step 1: Develop a cyber-physical collaborative correlation analysis and collaborative active power control architecture; Step 2, design a centralized control method based on power flow constraints at the upper layer; Step 3, designing a combined fault-tolerant control method at the lower layer; Step 4: Verify the effectiveness of this method by building an experimental scenario.

2. According to claim 1, a method for controlling active power of a DC distribution network with cyber-physical collaboration for multi-microgrid access, characterized in that: In step 1, information-physical collaboration includes the network layer and the physical layer; association analysis means that the physical layer knows the network situation and the network layer knows the needs of the physical business; cooperation means that the network layer should provide reliable communication support according to the business needs of the physical layer; the control method of the physical layer should be adaptively adjusted according to the support capabilities that the network layer can provide.

3. According to claim 1, a method for controlling active power of a DC distribution network with cyber-physical collaboration for access to multiple microgrids is characterized in that: In step 2, the upper-level design is designed from the perspectives of cyber-physical correlation analysis and collaboration. The specific design methods are as follows: Step 21, information-physical correlation analysis; Step 22: Cyber-physical collaboration and coordination.

4. According to claim 3, a method for controlling active power of a DC distribution network with cyber-physical collaboration for multi-microgrid access is characterized in that: The specific design method in step 21 is as follows: Step 211: The information layer analyzes the business requirements of the physical layer; Model the relationship between the execution effect of active power regulation and the transmission data; based on the sensitivity analysis method, quantify the impact of changes in data transmission speed and accuracy on the active power regulation effect to determine the importance of different data; when facing business G, when its relationship with related parameters is modeled as G=f(x1,…,x m ) (1) where x1,…,x m is a data set, and f is a mapping relationship; the impact of changes in the transmission speed and accuracy of the i-th parameter on the execution effect of the service G can be expressed by sensitivity, overshoot, and steady-state error; Step 212: The physical layer associates and analyzes the network environment that the information layer can provide; after network matching is completed, information about the router status, network architecture, and transmission channel status will be fed back to the physical layer; based on the above data, the physical layer adjusts the active power control strategy in real time; In this way, the control strategy can be adjusted accordingly.

5. According to claim 3, a method for controlling active power of a DC distribution network with cyber-physical collaboration for multi-microgrid access, characterized in that: Step 22: Cyber-physical collaboration and coordination; specifically includes the following: Step 221: Flow constraint-dependent centralized control; specifically: Step 2211: Optimizing model construction; Given m microgrids with power fluctuations, for the i-th microgrid, the control command is set as the multi-objective optimization model is Where ΔP α / P α and ΔP β / P β is the ratio of the active power change of the αth and βth microgrids to their capacity; Speed1 and Speed m is the response speed; △P α , △P β and △P m is the active power change; P loss.α is the line loss of the αth microgrid; △U α is the voltage fluctuation between the αth buses; || abs Indicates taking the absolute value; || Nor represents the normalized operation; ρ1+ρ2+ρ3+ρ4=1. The relevant constraint is power balancing: And adjustable capacity limit: 0≤△P α ≤△P α.max ; Step 2212: Optimizing model construction; Set constraints, response time constraint: max(△P1 / Speed 1Nor ,…,△P m / Speed mNor )≤T max (3) Line loss constraints: Node voltage limit: U α.min ≤U α ≤U α.max (5) where △P α is the output active power of the αth microgrid; △P α.max is the maximum adjustable power of the αth microgrid; △P α.loss is the line loss of the αth microgrid; P lineloss.max is the loss threshold set; U α.min and U α.max are the minimum and maximum voltages of the node; T max is the maximum response time set; Step 222: Demand-driven wired network matching, including Step 2221 optimizes the model construction: First, the sensitivity calculation should be completed in the active power control; assuming that there are m buses, the static equation of the sensor injected power change in the αth bus on the target (2) is in is the partial derivative of Z in (2) with respect to the output power of the αth microgrid; after determining the sensitivity, the flow data to be transmitted is arranged in descending or descending order of sensitivity, and we get In addition, the centralized tree network is planned according to the order in (7); the relevant objective function is constructed as Where Sign is the sign function; Per αβ is the performance of the channel from the αth router to the βth router; a αβ is the relationship between the αth and βth routers; if the αth router can send data to the βth router, a αβ =1 or a αβ =0; otherwise, N α is the set of neighbor routers of the αth router; Step 2221: Constraints: The constraints for uploading the network are constructed as follows: For the αth router, the number of input channels and output channels should not exceed the allowed values: and Number of connected channels >= Number of microgrids: Each transport router has its output channels and input channels: and Each microgrid has at least one output channel: The microgrid control center should be connected to at least the same number of input channels as the microgrid: For the download network, constraints (12) and (13) are modified to the following equations, while other constraints remain unchanged: Each MG has at least one input channel: The center has at least one down-load channel: Where Num out.α and Num in.α is the number of output and input channels allowed by the αth node; N Tr is a set of transmission routers; once the above optimization problem is reformulated, the paths can be obtained in sequence.

6. The method for controlling active power of a DC distribution network in cyber-physical collaboration for multi-microgrid access according to claim 1, characterized in that: In step 3, after completing the upper layer design, the design of the active power regulation strategy and dual path optimization strategy of the lower layer with active fault tolerance is carried out; Specifically, step 31 includes designing a combined fault-tolerant control method: According to the control command of the microgrid, based on the equal capacity ratio P h-α / P h-α.cap =P h-β / P h-β.cap , represents the ratio of the output command to the rated capacity of the α / β distributed resources and the control command of the "virtual leader-follower consistency control", adaptively adjusts the controller parameters to enable the resources to flexibly respond to the commands of the microgrid; the control equation for the performance of the αth resource is △P h-α =△P h-α.Ref (16) Where △P h-α and △P h-β is the active power adjustment of the αth and βth distributed resources in the hth microgrid; P h-α.cap and P h-β.cap is the remaining capacity of the αth and βth resources; △P h-α.Ref and △P h-β.Ref is the inverse solution of the control result, namely △P h-α.Ref =γ h-α (t)·P h-α.cap and △P h-β.Ref =γ h-β (t)·P h-β.cap obtained; in this case, γ is usually chosen h-L (t) is Where P MGh is the power required by the hth microgrid; the relevant performance equation of distributed resources is The output command is where u h-1 (t) and u h-n (t) is the controller; k h-α1 , k h-α2 , k h-n2 and k h-n1 is the gain; a h-αβ represents the correlation between the αth and βth resources; b h-α represents the association between the αth resource and the volumetric load; During regulation, the deviation between the controlled variable and the set value of the volume load is e h =u h -γ h-L ; Under normal circumstances, the controlled deviation is adjusted to 0 by selecting appropriate parameters; however, if the network switches, the disturbance will have a negative impact on the performance quality, and sliding mode control is needed to suppress it; the affected controller u h ' -1 (t)~u h ' -n The equation of (t) is (19), where △γ h-1 (t) and △γ h-n (t) is the introduced disturbance; The controllable error u′-u is Where △u h-1 (t) = u h ' -1 (t)-u h-1 (t); The feedback matrix F is used to complete the sliding mode control, which is designed as Among them, F h-αβ =-K γh-αβ a h-αβ ; F h-αα needs to be designed; taking the αth distributed resource as an example, the sliding surface is designed as S = K P △γ h-α +K I ∫△γ h-α dt, where and is the coefficient; the associated controller design is F h-αα =F h-αa +F h-αb ;make and F h-α1 =F h-αa , we can get where F h-αa is the solution of equation (21); then, the Lyapunov function is taken as L = S 2 / 2, and make it satisfy Lyapunov stability, considering F h-αα =F h-αa +F h-αb , we can get in is the switch gain in the protocol; Step 32: Design of dual path optimization strategy The design process is as follows: First, when the microgrid operates in an alarm state, the sensitivity between the controlled variables and the relevant data should be calculated; when the system operates in an emergency state, the sensitivity between the controlled variables and time should be calculated; taking the αth distributed resource as an example, the partial derivative between its controller and the data of the βth distributed resource is Its partial derivative with respect to time is In addition, the path needs to be updated; the upper-level optimization mainly solves the problem in the alert state, based on the principle of matching path performance with data importance, and reconstructs the directed spanning tree using global information; when the αth distributed resource needs to be connected to the directed tree, the transmission path should match the data in sequence; the optimization target adopts The constraints are: Number of channels DER: There is at most one connected channel between two routers: The controlled DER has only one input channel: Volumetric loads have at least one output channel and no input channels: Where P αβ is the security probability of the channel from the βth router to the αth router; Number DER is the number of distributed resources that need to be connected; The lower-layer optimization mainly solves the problem of path reconstruction in emergency situations. It is based on the transmission network constructed in the upper layer. First, the selected path is set to 1, and the unselectable path and the faulty path are set to 0. In addition, based on the principle of matching transmission speed and data input, distributed resources are used to quickly generate path reconstruction solutions using local information. The target model is where t αβ is the transmission time of the channel from the αth distributed resource to the βth distributed resource; if the i-th distributed resource needs to be connected to the network; the following constraints need to be considered in the optimization model; Transport security constraints: Ensure that the i-th distributed resource will be reconnected to the directed tree: Among them, Prob set is the required safety probability; if multiple sets of data need to be reconnected, the data should be matched to the channels in sequence.

7. A cyber-physical collaborative active power control device for a DC distribution network with multiple microgrids connected, comprising: The overall structural module is configured to: formulate information-physical collaborative correlation analysis and collaborative active power control architecture. The upper level design module is configured to design a centralized control method based on power flow constraints. The lower-level design module is configured to: design a combined fault-tolerant control method.

Images