Distributed Communication Resource Allocation and Voltage Coordination Regulation Method and Device

Abstract

The present application discloses a method and device for distributed communication resource allocation and voltage collaborative regulation, belonging to the technical field of power system operation optimization. The method includes: for a new power system, constructing a wireless network control system based on a 5G local area network; determining a wireless transmission model of the wireless network control system and a dynamic model of voltage control; based on the wireless transmission model and the dynamic model of voltage control, determining a cross-domain dependence coupling relationship between the wireless communication and control system performance of the wireless network control system; based on the cross-domain dependence coupling relationship, determining a cross-domain optimization problem; and using the primal-dual graph neural network algorithm to solve the cross-domain optimization problem to obtain an optimal voltage control and communication power allocation strategy.

Description

Technical Field

[0001] This application belongs to the technical field of power system operation optimization, and particularly relates to a method and device for distributed communication resource allocation and voltage coordinated control. Background Art

[0002] With the increase in the penetration rate of renewable energy in the distribution network of the new power system, the active and reactive power injected by distributed generators (DGs) into the main power grid often changes. The traditional centralized structure voltage control obtains the global information of DGs through a central calculation and communication unit. Under the characteristics of a large number of terminals and wide coverage in the distribution network, problems such as the failure of control information interaction caused by communication packet loss can easily lead to suboptimal control decisions. The 5G network can give play to its advantages of ultra-high bandwidth and low latency. However, it is difficult to achieve better overall performance of a multi-loop wireless network control system only through power management, and there is still a problem of insufficient service coordination, which cannot effectively meet the above-mentioned distribution network control service requirements. Summary of the Invention

[0003] This application aims to solve at least one of the technical problems existing in the related art. For this purpose, this application proposes a method and device for distributed communication resource allocation and voltage coordinated control, realizing the coordinated control of distributed communication resources and voltage.

[0004] In a first aspect, this application provides a method for distributed communication resource allocation and voltage coordinated control, which includes:

[0005] For the new power system, construct a wireless network control system based on a 5G local area network;

[0006] Determine the wireless transmission model and the dynamic model of voltage control of the wireless network control system;

[0007] Based on the wireless transmission model and the dynamic model of voltage control, determine the cross-domain dependence coupling relationship between the wireless communication and the control system performance of the wireless network control system;

[0008] Based on the cross-domain dependence coupling relationship, determine the cross-domain optimization problem;

[0009] Use the primal-dual graph neural network algorithm to solve the cross-domain optimization problem to obtain the optimal voltage control and communication power allocation strategy.

[0010] In the above technical scheme, for the new power system, a wireless network control system based on 5G local area network is constructed, the wireless transmission model of the wireless network control system and the dynamic model of voltage control are determined, and then based on the wireless transmission model and the dynamic model of voltage control, the dependency coupling relationship between the control domain and the communication domain is determined. Based on the cross-domain dependency coupling relationship, the cross-domain optimization problem is determined, and the original dual graph neural network algorithm is used to solve the cross-domain optimization problem to obtain the optimal voltage control and communication power allocation strategy, realizing distributed communication resource allocation and coordinated control of voltage, which can ensure that the voltage control system of the variable new energy distribution network has ideal dynamic and steady-state response characteristics and meets the business communication network requirements.

[0011] According to one embodiment of the present application, for the new power system, a wireless network control system based on a 5G local area network is constructed, including:

[0012] The distributed controller is connected to multiple distribution network nodes of the new power system through the 5G local area network to realize distributed communication and control, and dynamically adjust the voltage control and communication power allocation strategy in each control cycle to cope with voltage fluctuations and load changes in the distribution network.

[0013] In the above technical solution, the distributed controller is connected to multiple distribution network nodes of the new power system through the 5G local area network, realizing distributed communication and control. The voltage control strategy and communication power allocation strategy can be dynamically adjusted in each control cycle, which can effectively cope with voltage fluctuations and load changes in the distribution network, and improve the control performance and communication performance of the new power system.

[0014] According to one embodiment of the present application, the connecting of the distributed controller to multiple distribution network nodes of the novel power system through a 5G local area network includes:

[0015] Determining a plurality of control areas according to a distribution network topology structure corresponding to the novel power system;

[0016] Wireless bearer is realized between different control areas through network nodes;

[0017] In each control area, the network node covers at least one distributed energy node, and the distributed energy node includes: a new energy generation node or a common distributed generation node. The common distributed generation node serves as the distribution network node and is sensed by the associated new energy generation node. The common distributed generation node and the associated new energy generation node communicate through an end-to-end link in the control area to which they belong;

[0018] The status information of the ordinary distributed generation nodes is transmitted to the new energy generation nodes through the end-to-end link, forming the upstream status perception of the control loop. The new energy generation nodes, as the distributed controllers, aggregate the upstream status information through the end-to-end link, determine the reactive power injection ratio value, and execute the downstream voltage control command.

[0019] The network node distributes power for the end-to-end link within the control area covered by the network node and transmits the power distribution instruction to the new energy generation node.

[0020] In the above technical solution, the status information of the ordinary distributed generation nodes is transmitted to the new energy generation nodes through the end-to-end link, forming the upstream status perception of the control loop. The new energy generation nodes, as the distributed controllers, aggregate the upstream status information through the end-to-end link, determine the reactive power injection ratio value, and execute the downstream voltage control command, thereby realizing voltage control. The network node distributes power for the end-to-end link within the covered control area and transmits the power distribution instruction to the new energy generation node, so that the new energy generation node can communicate with the ordinary distributed generation nodes associated with it according to the allocated communication power, realizing the conversion from the power system to the wireless network control system.

[0021] According to an embodiment of the present application, the m-th new energy generation node associated with the adjacent ordinary distributed generation nodes is set as the m-th distributed controller of the wireless network control system, forming the m-th control subsystem. Multiple such control subsystems form a closed feedback loop through the 5G local area network.

[0022] Wherein, the state of the m-th new energy generation node at a certain moment depends on its own measurement value and the measurement values obtained from its neighbors.

[0023] In the above technical solution, the m-th new energy generation node associated with the adjacent ordinary distributed generation nodes is set as the m-th distributed controller of the wireless network control system, forming the m-th control subsystem. A closed feedback loop is formed among multiple control subsystems through the 5G local area network, and sensing information and control command information are transmitted through the wireless medium, realizing a wireless network voltage control system based on the 5G local area network, and further enabling the collaborative optimization between the control system and the communication system.

[0024] According to an embodiment of the present application, determining the cross-domain dependence coupling relationship between the wireless communication and control system performance of the wireless network control system based on the wireless transmission model and the dynamic model of voltage control includes:

[0025] Based on the Lyapunov stability theory and the wireless transmission model, perform stability and rapidity analysis on the dynamic model of the voltage control to determine the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions.

[0026] In the above technical solution, based on the Lyapunov stability theory and the wireless transmission model, perform stability and rapidity analysis on the dynamic model of the voltage control to determine the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions, and reflect the cross-domain dependent coupling relationship between the performance of wireless communication and control systems through the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions. Subsequently, determine the cross-domain optimization problem based on the cross-domain dependent coupling relationship, which can ensure that the wireless network voltage control closed-loop system has ideal dynamic and steady-state response characteristics while meeting communication requirements.

[0027] According to an embodiment of the present application, determining the cross-domain optimization problem based on the cross-domain dependent coupling relationship includes:

[0028] Determine the objective function, where the objective function includes: a first objective function for making the corresponding voltage converge to the desired range within a preset duration, and a second objective function for introducing penalties for voltage deviation and reactive power injection;

[0029] According to the dynamic model rules, transient response requirements, the exponential stability constraint conditions and convergence rate constraint conditions under the imperfect wireless link conditions, transmit power constraint conditions, and packet loss rate constraint conditions, determine the objective constraint conditions;

[0030] According to the objective function and the objective constraint conditions, determine the cross-domain optimization problem.

[0031] In the above technical problem, the cross-domain optimization problem aims to make the corresponding voltage converge to the desired range within a preset duration and introduce penalties for voltage deviation and reactive power injection. Considering constraints such as dynamic model rules, transient response requirements, exponential stability and convergence rate under imperfect wireless link conditions, transmit power, and packet loss rate, it realizes distributed communication resource allocation and coordinated control of voltage, which can ensure that the voltage control system of the variable renewable energy distribution network has ideal dynamic and steady-state response characteristics and meets the requirements of the service communication network.

[0032] According to an embodiment of the present application, solving the cross-domain optimization problem using the primal-dual graph neural network algorithm to obtain the optimal voltage control and communication power allocation strategy includes:

[0033] Parameterize the constraint conditions in the cross - domain optimization problem to establish the constraint conditions in the design of the graph neural network model, forming a parameterized joint optimization problem;

[0034] Express the parameterized joint optimization problem through the Lagrangian dual function that simultaneously considers the optimization objective and the constraint conditions;

[0035] Model the wireless network control system as a directed communication graph, and determine the node features and edge features of the directed communication graph;

[0036] Input the node features and edge features into the graph neural network model, use the Lagrangian dual function as the loss function, and combine primal - dual learning to train the graph neural network model. After the training is completed, obtain the output of the graph neural network model to get the optimal voltage control and communication power allocation strategy.

[0037] In the above technical solution, parameterize the constraint conditions in the cross - domain optimization problem to establish the constraint conditions in the design of the graph neural network model, forming a parameterized joint optimization problem; express the parameterized joint optimization problem through the Lagrangian dual function that simultaneously considers the optimization objective and the constraint conditions, model the wireless network control system as a directed communication graph, determine the node features and edge features of the directed communication graph, input the node features and edge features into the graph neural network model, use the Lagrangian dual function as the loss function, and combine primal - dual learning to train the graph neural network model. After the training is completed, obtain the output of the graph neural network model to get the optimal voltage control and communication power allocation strategy, which can provide a design for the information interaction and transmission scheme for remote control under the condition of high new - energy access ratio in the distribution network, consider the requirements of the control system and the limitations of spectrum and energy resources, obtain the optimal voltage control strategy and communication resource allocation strategy, realize the collaborative control of distributed communication resource allocation and voltage, and improve the information exchange ability and control performance of the new power system.

[0038] In a second aspect, the present application provides a distributed communication resource allocation and voltage collaborative regulation device, and the device includes:

[0039] A system construction unit, configured to construct a wireless network control system based on a 5G local area network for the new power system;

[0040] A model determination unit, configured to determine the wireless transmission model and the dynamic model of voltage control of the wireless network control system;

[0041] A cross - domain dependency determination unit, configured to determine the cross - domain dependency coupling relationship between the wireless communication and the control system performance of the wireless network control system based on the wireless transmission model and the dynamic model of voltage control;

[0042] An optimization problem determination unit for determining a cross-domain optimization problem based on the cross-domain dependency coupling relationship;

[0043] A solving unit for solving the cross-domain optimization problem by using the primal-dual graph neural network algorithm to obtain an optimal voltage control and communication power allocation strategy.

[0044] In the above technical solution, for the new power system, a wireless network control system based on a 5G local area network is constructed, the wireless transmission model of the wireless network control system and the dynamic model of voltage control are determined, and then based on the wireless transmission model and the dynamic model of voltage control, the dependency coupling relationship between the control domain and the communication domain is determined. Based on the cross-domain dependency coupling relationship, a cross-domain optimization problem is determined, and the primal-dual graph neural network algorithm is used to solve the cross-domain optimization problem to obtain an optimal voltage control and communication power allocation strategy, realizing distributed communication resource allocation and coordinated control of voltage, which can ensure that the voltage control system of the variable new energy distribution network has ideal dynamic and steady-state response characteristics and meets the requirements of the service communication network.

[0045] In a third aspect, the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the distributed communication resource allocation and voltage coordinated regulation method as described in the first aspect above.

[0046] In a fourth aspect, the present application provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the distributed communication resource allocation and voltage coordinated regulation method as described in the first aspect above.

[0047] In a fifth aspect, the present application provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the distributed communication resource allocation and voltage coordinated regulation method as described in the first aspect.

[0048] In a sixth aspect, the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the distributed communication resource allocation and voltage coordinated regulation method as described in the first aspect above.

[0049] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. Description of the Drawings

[0050] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the description of the embodiments in conjunction with the following drawings, where:

[0051] Figure 1 FIG. 4 is one of the flow diagrams of the distributed communication resource allocation and voltage co-regulation method provided by the embodiments of the present application;

[0052] Figure 2 FIG. 8 is a schematic diagram of the architecture of a wireless network control system based on a 5G local area network provided by the embodiments of the present application;

[0053] Figure 3 FIG. 12 is a flow diagram of using the primal-dual graph neural network algorithm to solve the cross-domain optimization problem to obtain the optimal voltage control and communication power allocation strategy provided by the embodiments of the present application;

[0054] Figure 4 FIG. 16 is a schematic diagram of the structure of the distributed communication resource allocation and voltage co-regulation device provided by the embodiments of the present application;

[0055] Figure 5 FIG. 20 is a schematic diagram of the structure of an electronic device provided by the embodiments of the present application. Detailed Embodiments

[0056] The technical solutions in the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application belong to the scope of protection of the present application.

[0057] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data may be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order different from those illustrated or described herein, and the objects distinguished by "first", "second", etc. generally belong to the same category, and the number of objects is not limited. For example, the first object may be one or more. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally indicates an "or" relationship between the associated objects before and after.

[0058] The distributed communication resource allocation and voltage co-regulation method and device provided by the embodiments of the present application will be described in detail below with reference to the drawings, through specific embodiments and their application scenarios.

[0059] The execution subject of the distributed communication resource allocation and voltage co-regulation method provided by the embodiments of this application can be an electronic device or a functional module or functional entity in the electronic device that can implement the distributed communication resource allocation and voltage co-regulation method. The electronic devices mentioned in the embodiments of this application include, but are not limited to, processors, controllers, computing platforms, cloud platforms, servers, terminals, etc. Hereinafter, taking the electronic device as the execution subject as an example, the distributed communication resource allocation and voltage co-regulation method provided by the embodiments of this application will be described.

[0060] As Figure 1 shown, the distributed communication resource allocation and voltage co-regulation method includes: Step 110, Step 120, Step 130, Step 140, and Step 150.

[0061] Step 110: For the new power system, construct a wireless network control system based on a 5G local area network;

[0062] Step 120: Determine the wireless transmission model and the dynamic model of voltage control of the wireless network control system;

[0063] Step 130: Based on the wireless transmission model and the dynamic model of voltage control, determine the cross-domain dependent coupling relationship between the wireless communication and the control system performance of the wireless network control system;

[0064] Step 140: Based on the cross-domain dependent coupling relationship, determine the cross-domain optimization problem;

[0065] Step 150: Use the primal-dual graph neural network algorithm to solve the cross-domain optimization problem to obtain the optimal voltage control and communication power allocation strategy.

[0066] The new power system is gradually evolving into a large-scale time-varying topology with a large number of distributed generations, rooftop photovoltaic (PV) penetrations, and explosive information exchanges. Distributed energy resources, especially numerous geographically dispersed distributed photovoltaic systems, introduce significant uncertainties in critical production control operations, such as photovoltaic regulation and new load control. Information exchange in the control loop of wireless communication is often vulnerable to the influence of transmission packet loss probability and uncertain network-induced delay. Dynamic wireless links severely limit control performance, and it is particularly important to improve the reliable transmission ability during the sharing of available wireless media.

[0067] For wireless applications in the power industry, the bandwidth of the dedicated 230 MHz frequency band is only 1 MHz, which cannot meet the requirements of typical key control scenarios including intelligent distributed automation and distributed energy regulation. The current wireless communication network used in the power system lacks an effective and reliable communication mechanism to promote the reliable and economical access of numerous distributed control terminals. This limitation hinders the realization of comprehensive network coverage, prevents the ability to meet the needs of distributed terminals, and cannot meet the requirements of on-site high-frequency and real-time data communication. Compared with the 230 MHz dedicated network and the 5G public network, the 5G Local Area Network (LAN) wireless communication technology emphasizes Ultra-Reliable Low-Latency Communications (URLLC), which closely matches the requirements of the power system and can meet the communication needs of universal, flexible, and mobile access without relying on physical cables. Therefore, by using the 5G LAN wireless communication technology, the terminal sensors and actuators of the power control system can be widely deployed to monitor and control physical processes, and numerous small-scale integrated control subsystems sharing 5G communication resources form a wireless network control system to transmit sensing information and control command information through the wireless medium.

[0068] Related technologies only focus on optimizing control design to adapt to the physical system state and assume that the packet loss rate on the unreliable wireless channel has been fully described. The clear relationship between control performance, transmission reliability, and time delay remains unclear, which makes the analysis and optimization of control performance somewhat challenging.

[0069] This application aims at the goal of reliable transmission and deterministic guarantee for wireless bearers of power distributed control services, and proposes a method for distributed communication resource allocation and voltage coordinated regulation.

[0070] This application first constructs a wireless network control system based on the 5G local area network for the new power system, that is, based on the physical topology of the new power system, a wireless network control system based on the 5G local area network can be constructed. The physical topology of the new power system can be divided into multiple control regions, and the control regions are divided based on the main branch nodes of the physical topology.

[0071] Then, considering the reliability of wireless communication, packet loss rate, etc., the wireless transmission model of the wireless network control system is determined, and considering the impact of the unreliability of wireless communication on voltage control, the dynamic model of voltage control of the wireless network control system is determined.

[0072] Further, considering the stability and rapidity of the wireless network control system, based on the wireless transmission model and the dynamic model of voltage control, the cross-domain dependence coupling relationship between the wireless communication and the control system performance of the wireless network control system is determined, where the cross-domain dependence coupling relationship refers to the dependence coupling relationship between the control domain and the communication domain.

[0073] Further, considering the characteristics of the variable topology structure of new energy and the cross-domain dependence coupling relationship, under the premise of meeting the requirements of the control system and the spectrum resource constraints, the cross-domain optimization problem is determined, and then the voltage control and the communication power allocation are jointly optimized.

[0074] Finally, the original-dual graph neural network algorithm is used to obtain the optimal voltage control strategy and the communication power allocation strategy, ensuring that the variable new energy distribution network voltage control system has ideal dynamic and steady-state response characteristics and meets the requirements of the service communication network.

[0075] The distributed communication resource allocation and voltage collaborative regulation method provided by the embodiments of the present application constructs a wireless network control system based on a 5G local area network for a new power system, determines the wireless transmission model and the dynamic model of voltage control of the wireless network control system, and then based on the wireless transmission model and the dynamic model of voltage control, determines the dependence coupling relationship between the control domain and the communication domain. Based on the cross-domain dependence coupling relationship, the cross-domain optimization problem is determined, and the original-dual graph neural network algorithm is used to solve the cross-domain optimization problem to obtain the optimal voltage control and the communication power allocation strategy, realizing the distributed communication resource allocation and the collaborative control of voltage, and ensuring that the variable new energy distribution network voltage control system has ideal dynamic and steady-state response characteristics and meets the requirements of the service communication network.

[0076] In some embodiments, step 110 constructs a wireless network control system based on a 5G local area network for a new power system, including:

[0077] Connect the distributed controller to multiple distribution network nodes of the new power system through a 5G local area network for realizing distributed communication and control, and dynamically adjust the voltage control and the communication power allocation strategy within each control cycle to cope with voltage fluctuations and load changes in the distribution network.

[0078] Distribution network nodes usually include power generation nodes, power transmission and transformation nodes, energy storage nodes, load nodes, etc. In the embodiments of the present application, the distribution network nodes are power generation nodes.

[0079] The embodiment of the present application takes into account the variable topology characteristics of new energy sources, and connects a distributed controller to multiple distribution network nodes of a new power system through a 5G local area network, wherein the distributed controller is deployed on or near a new energy power generation node (such as a photovoltaic (PV) node). The new energy power generation node can be used as a distributed controller, and the ordinary distributed power generation node can be used as a distribution network node. One new energy power generation node can be associated with multiple ordinary distributed power generation nodes, and the new energy power generation node and the ordinary distributed power generation node communicate through a D2D link, thereby realizing distributed communication and control, and dynamically adjusting the voltage control strategy and the communication power allocation strategy in each control cycle to cope with voltage fluctuations and load changes in the distribution network.

[0080] The distributed communication resource allocation and voltage coordinated regulation method provided in the embodiment of the present application connects the distributed controller to multiple distribution network nodes of the new power system through the 5G local area network, realizes distributed communication and control, and can dynamically adjust the voltage control strategy and communication power allocation strategy in each control cycle, which can effectively cope with voltage fluctuations and load changes in the distribution network, and improve the control performance and communication performance of the new power system.

[0081] In some embodiments, the connecting of the distributed controller to multiple distribution network nodes of the novel power system through a 5G local area network includes:

[0082] Determining a plurality of control areas according to a distribution network topology structure corresponding to the novel power system;

[0083] Wireless bearer is realized between different control areas through network nodes;

[0084] In each control area, the network node covers at least one distributed energy node, and the distributed energy node includes: a new energy generation node or a common distributed generation node. The common distributed generation node serves as the distribution network node and is sensed by the associated new energy generation node. The common distributed generation node and the associated new energy generation node communicate through an end-to-end link in the control area to which they belong;

[0085] The state information of the common distributed power generation node is transmitted to the new energy power generation node through the end-to-end link to form an uplink state perception of the control loop. The new energy power generation node acts as the distributed controller to aggregate the uplink state information through the end-to-end link, determine the reactive power injection ratio value, and execute the downlink voltage control command;

[0086] The network node performs power allocation for end-to-end links within a control area covered by the network node, and transmits a power allocation instruction to the new energy power generation node.

[0087] Figure 2 This is a schematic diagram of the architecture of a wireless network control system based on a 5G local area network provided in an embodiment of the present application. Figure 2 The wireless network control system in the embodiment of the present application is a wireless network voltage control system (or a wireless network voltage control closed-loop system) that takes into account high-proportion photovoltaic injection. Multiple control areas can be determined based on the distribution network topology corresponding to the new power system, specifically, by dividing the control area based on the main branch nodes in the distribution network topology.

[0088] In each control area, the network node covers at least one distributed energy node. It should be noted that the network node refers to a node that provides network services, such as a base station. Optionally, each control area includes a new energy power generation node (such as a photovoltaic (PV) node) and at least one ordinary distributed generation (DG) node, and the network node covers a new energy power generation node in the corresponding control area and at least one ordinary distributed generation node associated with the new energy power generation node. Among them, the ordinary distributed generation node, as the distribution network node, is perceived by the new energy power generation node associated with the ordinary distributed generation node, and the ordinary distributed generation node and the associated new energy power generation node communicate through an end-to-end link in the control area to which they belong.

[0089] The state information of the common distributed power generation node is transmitted to the new energy power generation node through the end-to-end link to form an uplink state perception of the control loop. The new energy power generation node, as the distributed controller, aggregates the uplink state information through the end-to-end link, determines the reactive power injection ratio value, and executes the downlink voltage control command to adjust the reactive power according to the reactive power injection ratio value, thereby realizing voltage control;

[0090] The network node distributes power for the end-to-end links within the covered control area, and transmits the power distribution instruction to the new energy power generation node, so that the new energy power generation node can communicate with the ordinary distributed power generation node associated with it according to the allocated communication power.

[0091] refer to Figure 2 The wireless network control system in the embodiment of the present application includes a cellular base station in each control area covering at least one distributed energy source (DER) node, and these DER nodes are classified into DG nodes and PV nodes.

[0092] The wireless communication network under consideration is In the region, Base stations in the same frequency bandwidth Cover PV nodes and the remaining DGs. The set of DGs and PVs is divided into groups, and the set of DG indices in the th group is , and these DGs can be sensed by all associated PVs. The set of PVs in the th area is , indicating the indices of these PV nodes in the total set . Considering a wireless network carried by a 5G local area network type service, where PV-DG transceiver pairs communicate through a device-to-device (D2D, also known as end-to-end) mode within the area, The set of communication links of PV nodes is , and the set of all D2D links is denoted as

[0093] The distributed communication resource allocation and voltage coordinated regulation method provided by the embodiments of this application. The status information of ordinary distributed generation nodes is transmitted to new energy generation nodes through end-to-end links to form the upstream status perception of the control loop. The new energy generation nodes, as distributed controllers, aggregate the upstream status information through end-to-end links, determine the reactive power injection ratio value, and execute the downstream voltage control command, thereby realizing voltage control. The network node allocates power for the end-to-end links within the covered control area and transmits the power allocation instruction to the new energy generation nodes, so that the new energy generation nodes can communicate with the ordinary distributed generation nodes associated with them according to the allocated communication power, realizing the conversion from the power system to the wireless network control system.

[0094] In some embodiments, the mth new energy generation node associated with adjacent ordinary distributed generation nodes is set as the mth distributed controller of the wireless network control system to form the mth control subsystem, and multiple such control subsystems form a closed feedback loop through the 5G local area network;

[0095] wherein, the status of the mth new energy generation node at a certain moment depends on its own measurement value and the measurement values obtained from its neighbors.

[0096] It is understandable that the wireless network control system is composed of multiple control subsystems, and a closed feedback loop is formed among the multiple control subsystems through the 5G local area network.

[0097] The distributed communication resource allocation and voltage coordinated regulation method provided by the embodiment of the present application sets the mth new energy generation node associated with the adjacent ordinary distributed generation node as the mth distributed controller of the wireless network control system, forms the mth control subsystem, and a closed feedback loop is formed among the multiple control subsystems through the 5G local area network. Sensing information and control command information are transmitted through the wireless medium, realizing a wireless network voltage control system based on the 5G local area network, and further realizing the collaborative optimization between the control system and the communication system.

[0098] In some embodiments, step 120 of determining the wireless transmission model and the dynamic model of voltage control of the wireless network control system includes:

[0099] 1) Determine the wireless transmission model

[0100] a) Channel model

[0101] Considering that in each frame, the DG sends information bits to its associated PV. In the frame, use to represent the channel coefficient of the direct communication link , use to represent the channel coefficient of the interference link between links and ( ), is the number of DGs in the kth control area, and these coefficients are affected by small-scale fading and large-scale attenuation. Therefore, the channel coefficient matrix of the th PV is defined as

[0102]

[0103] where represents the path loss factor, represents the square root of the large-scale channel gain, and its elements are respectively denoted as and , represents the small-scale channel gain, represents the successive product of elements.

[0104] Denote the instantaneous power matrix of PVs as , where represents the instantaneous power used in the link , Indicates the instantaneous power used in the interfering link between the link and . Therefore, the signal-to-noise-plus-interference ratio (SINR) of link is expressed as

[0105]

[0106] where represents the noise power.

[0107] b) Short-packet model

[0108] The status information of DG is transmitted to PV through the D2D link to form the uplink status perception of the control loop, while the distributed PV executes the downlink control command locally. Therefore, when the delay exceeds the maximum delay threshold, the packet loss rate in the uplink D2D transmission is determined. Assume that the total bandwidth of the system is , and the bandwidth allocated to link is . The transmission rate of the communication channel of the D2D link can be expressed as

[0109]

[0110] The status information transmitted in the control system is mainly short data packets. Assume that the size of the transmitted data is b bits, the sampling interval is , and bits of data are generated in each time slot. The status arrival process can be regarded as a queue. Subsequently, the concept of effective capacity is used to derive the constraint conditions for ensuring the queuing delay requirement and the packet loss probability. Let the arrival rate of the th uplink queue of the th PV node be , and is the number of PV nodes in the kth control area. The effective capacity is , where represents the decay rate of the queue overflow probability, which is a constant in the control environment. The effective capacity can be approximated as

[0111]

[0112] Given as the delay threshold of the transmission delay , , the packet loss rate is expressed as

[0113]

[0114] 2) Determine the voltage control dynamic model

[0115] For the PV distributed control of the distribution network, the controller is close to the PV nodes and far from the widely distributed DG nodes. Assume that at each moment the -th PV node state depends only on its own measurements and the measurements obtained from its neighbors. The voltage of the -th PV node is

[0116]

[0117] where is the voltage of the -th adjacent DG, is the square of the current, and represent the active power and reactive power respectively, and represent the line resistance and reactance respectively. By setting the quadratic terms can be ignored, and an approximate model can be obtained and rewritten in vector form

[0118]

[0119] where and are vectors of length , is the voltage of the bus collected from the main grid through the smart distribution network. represents the uncontrollable voltage component determined by the active power injection of the load and PV, and are vectors of length .

[0120] The -th PV associated with the neighboring DG is represented as the -th control subsystem, whose state is , and the controlled variable is . The distributed controller determines the new reactive power injection value, thus triggering the voltage update, ensuring that the PV bus voltage is maintained within the tight range under any given operating conditions through . The closed-loop voltage control dynamics are described as follows:

[0121]

[0122]

[0123] In the case of applying state feedback control, the controller gain matrix is set to , that is , where is the control signal input, is the rated voltage under steady state, represents the rated voltage in vector form, the voltage vector of the m-th control subsystem, is the reactive power vector.

[0124] Multiple controllable subsystems close their feedback loops through a shared 5G wireless network, and the probability depends on the communication power allocated to the control subsystem and the random channel state of wireless fading. Due to limited wireless channels and significant electromagnetic interference, the information exchange in the control loop may not successfully reach the PV, thus introducing a packet loss rate that characterizes the reliability of the communication link to reflect the impact of the wireless network on control performance.

[0125] Given the packet loss rate and the binary random variable of Bernoulli distribution , when the th control subsystem's link 's uplink data packet is successfully received by the PV at time, and , is the packet loss rate threshold. Otherwise, under the influence of communication conditions, available spectrum resources, and transmission power, and . The control command actually executed by the th control subsystem is , is the voltage of the th PV node, = . Therefore, the control dynamics equation can be rewritten as,

[0126]

[0127] The distributed communication resource allocation and voltage collaborative regulation method provided by the embodiments of the present application determines the channel model and short packet model, calculates the communication conditions of the D2D link, determines the dynamic model of voltage control by considering the unreliability of communication, and subsequent cross-domain dependence coupling relationship analysis can be realized based on the wireless transmission model and the dynamic model of voltage control.

[0128] In some embodiments, step 130 determines the cross-domain dependence coupling relationship between the wireless communication of the wireless network control system and the control system performance based on the wireless transmission model and the dynamic model of voltage control, including:

[0129] Based on the Lyapunov stability theory and the wireless transmission model, perform stability and rapidity analysis on the dynamic model of the voltage control to determine the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions.

[0130] It can be understood that considering the impact of wireless communication on the performance of the control system, based on the Lyapunov stability theory and the wireless transmission model, perform stability and rapidity analysis on the dynamic model of the voltage control to determine the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions, and reflect the cross-domain dependent coupling relationship between wireless communication and the performance of the control system through the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions.

[0131] 1) Stability of the control system

[0132] In the voltage regulation task, the goal is to find a distributed controller that can drive the system state to finally reach the equilibrium point (i.e., the desired state). The embodiment of the present application designs a stable distributed controller based on the Lyapunov stability theory. For voltage control, the closed-loop dynamics can be expressed as , where represents the control input considering the reliability of the wireless link, represents the mapping function from the state to the control signal, represents the mathematical symbol that the voltage control equation is defined as . Consider the following Lyapunov function,

[0133]

[0134] where is a positive definite matrix, and is positive definite and radially unbounded. Based on the LaSalle invariance theorem, when and only when holds, indicating that for each initial voltage value, will converge to the largest invariant set , that is, the voltage trajectory returns to the acceptable upper and lower bound ranges.

[0135] Assume that for all adjacent DERs, is a continuously differentiable function, and when , . In addition, each satisfies and on

[0136]

[0137] and , then the dynamic system is asymptotically stable, where represents the packet loss rate vector.

[0138] In the case of achieving exponential stability, the Lyapunov difference has a decay rate , and its stability condition is

[0139]

[0140] 2) Rapidity of the control system

[0141] To consider the unpredictable impact of wireless communication on the control convergence performance, this application establishes the definition of the average dwell time. Given a mapping depending on the delay and switching signal caused by unreliability , where let be the sampling interval, represents the activation time interval of represents the packet loss time interval. For any , let represent the number of signal switches in the interval . If there exist and satisfying , then is the average dwell time of the switching signal, is the jitter boundary. If the closed-loop wireless network control system is exponentially stable with a decay rate of , where is a given positive constant, and holds for any control dynamics, is the set of integers, and at the same time .

[0142] Compared with the original convergence rate , the exponential decay rate increases by a factor of , indicating a decrease in the convergence performance caused by time-varying delays. The Lyapunov decay rate increases monotonically with and decreases monotonically with , where a larger represents fewer switching times, resulting in a smaller Lyapunov decay rate. The time-variation of the delay triggers the switching between subsystems, and the average dwell time needs to satisfy the lower bound to maintain stability. Therefore, if the change rate of the delay is small, the state of the closed-loop wireless network control system will converge to the nominal point at a faster speed. In this case, , that is, .

[0143] Based on the quadratic-like Lyapunov function , its expected inequality can be , let , the lower bound of the exponential convergence rate can be derived,

[0144]

[0145] Then the convergence rate of the control system needs to satisfy , is the transmission success rate of the th control subsystem at time t.

[0146] The distributed communication resource allocation and voltage cooperative regulation method provided by the embodiments of the present application, based on the Lyapunov stability theory and the wireless transmission model, performs stability and rapidity analysis on the dynamic model of the voltage control, determines the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions, reflects the cross-domain dependence coupling relationship between wireless communication and control system performance through the exponential stability constraint conditions and convergence rate constraint conditions under imperfect wireless link conditions, and subsequently determines the cross-domain optimization problem based on the cross-domain dependence coupling relationship, which can ensure that the wireless network voltage control closed-loop system has ideal dynamic and steady-state response characteristics while meeting communication requirements.

[0147] In some embodiments, step 140 determines the cross-domain optimization problem based on the cross-domain dependence coupling relationship, including:

[0148] Determine the objective function, where the objective function includes: a first objective function for making the corresponding voltage converge to the desired range within a preset time period, and a second objective function for introducing penalties for voltage deviation and reactive power injection;

[0149] Determine the objective constraint conditions according to the dynamic model rules, transient response requirements, the exponential stability constraint conditions and convergence rate constraint conditions under the imperfect wireless link conditions, transmission power constraint conditions, and packet loss rate constraint conditions;

[0150] Determine the cross-domain optimization problem according to the objective function and the objective constraint conditions.

[0151] The embodiments of the present application introduce transient and steady-state optimization problems in the control domain, aiming to achieve two goals: a) the voltage converges rapidly to the safe range around the nominal value ; b) maintaining a relatively economical steady-state cost.

[0152] Among them, the steady-state cost is defined as the cost of control actions, and the compact form is

[0153]

[0154] Among them, is a diagonal matrix , and , respectively representing the reactive power cost and the apparent power limit of each DER. It should be noted that the last term in the above equation is a constant. The above cost function can also be understood as a penalty for voltage deviation and reactive power injection.

[0155] Next, during the transient period [0, , the voltage needs to converge to the desired range, that is, . Combining the transient performance after voltage perturbation with the steady-state cost to ensure the closed-loop control performance. Therefore, the control optimization problem of the th control area is

[0156]

[0157]

[0158]

[0159]

[0160]

[0161]

[0162] Among them, the constraints and represent the dynamic model rules, represents the transient response requirement, and are the exponential stability constraint condition and the convergence rate constraint condition under imperfect wireless link conditions (to achieve the stability and rapidity of the control system).

[0163] Based on the effective capacity, considering the uplink energy efficiency index of the kth control area, it can be expressed as

[0164]

[0165] Among them, represents the circuit power consumption inevitable for the th DG due to electronic operation, represents the transmit power in the link , satisfying , represents the base station in the th control area The maximum transmission power. Therefore, the optimization problem can be formulated as maximizing the total energy efficiency,

[0166]

[0167]

[0168]

[0169]

[0170]

[0171] Among them, the constraint conditions Ensure that the upper limit of the total transmission power allocated within the th control area does not exceed the maximum value Ensure that the transmission power of each communication link is non - negative. The constraint condition Represents the packet loss rate in the th control loop Cannot exceed the reliability threshold And is subject to the coupling inequalities And Constraints.

[0172] The distributed communication resource allocation and voltage collaborative regulation method provided by the embodiments of the present application solves the cross - domain optimization problem so that the corresponding voltage converges to the desired range within a preset time, and introduces penalties for voltage deviation and reactive power injection as the goal, considering constraints such as dynamic model rules, transient response requirements, exponential stability and convergence rate under imperfect wireless link conditions, transmission power, packet loss rate, etc., realizing the collaborative control of distributed communication resources and voltage, and can ensure that the voltage control system of the variable new - energy distribution network has ideal dynamic and steady - state response characteristics and meets the requirements of the service communication network.

[0173] In some embodiments, as Figure 3 shown, step 150 uses the primal - dual graph neural network algorithm to solve the cross - domain optimization problem, obtaining the optimal voltage control and communication power allocation strategy, including:

[0174] Step 1501: Parameterize the constraint conditions in the cross - domain optimization problem to establish the constraint conditions in the design of the graph neural network model, forming a parameterized joint optimization problem;

[0175] The optimal voltage control problem requires explicit system dynamic knowledge, and network parameters are usually difficult to estimate, which prompts the present application to parameterize the original optimization problem. Specifically, the present application parameterizes the controller of each photovoltaic as , where the weights of the neural network are , is the control gain, = , therefore, . The constraint conditions and The control commands in the wireless network control system dynamics described depend on the packet loss rate and the voltage amplitude of distributed energy, where The controllable reactive power and uncontrollable active power in are collected into the photovoltaic. This application restricts the power allocation and control injection strategy to be Markovian, reflecting the dynamic conversion constraints.

[0176] The parameterization of the policy network needs to consider the constraint conditions , which restricts that the class of stable distributed controllers must be a strictly monotonically decreasing function. This application adopts a neural network design that satisfies the stability constraints, and constructs the stacked ReLU (SReLU) activation function as follows:

[0177]

[0178]

[0179] It is monotonically decreasing at and zero at ; meanwhile, it is monotonically decreasing at and zero at . Where and are weight vectors, and are bias vectors, combining hidden layer dimensions, . The variable form is .

[0180] Furthermore, this application achieves bounded by directly thresholding the slope, and the PV controller neural network parameterized controller is constructed as:

[0181]

[0182] Where

[0183]

[0184]

[0185] For the power allocation strategy with finite-dimensional parameters , it maps the channel state information of the D2D link to the normalized transmit power for uplink transmission, that is , where represents the channel gain matrix of the m-th control subsystem. Therefore, let and be the weight parameters, and The parameterized joint optimization problem can be reformulated as:

[0186]

[0187]

[0188]

[0189] Step 1502, express the parameterized joint optimization problem through the Lagrangian dual function that simultaneously considers the optimization objective and the constraint conditions;

[0190] The parameterization in the strategy provides significant computational and practical advantages. However, the constraints of the unsolved problem , especially the part involving cross-domain coupling inequalities. Therefore, considering the non-negative dual variables , corresponding to each constraint in the constraint conditions, the Lagrangian dual problem is derived. The Lagrangian function is defined as the following expression of the original problem:

[0191] -

[0192] Given the coupled Lagrangian function, the corresponding Lagrangian dual function can be defined as . This can be rewritten as a minimax optimization problem as follows:

[0193]

[0194] Step 1503, model the wireless network control system as a directed communication graph, and determine the node features and edge features of the directed communication graph;

[0195] Due to the complex cross-domain coupling and transient stability within a finite time, this application adopts the Stability-constrained Primal-dual GNN Algorithm (SPGNN) to effectively solve the constructed multi-objective problem, and processes the large-scale state-action space through the stability-constrained strategy parameterization.

[0196] To characterize the interaction between the PV and its neighboring distributed generators DG, the wireless network is modeled as a directed communication graph , where denote a set of end - to - end (D2D) communications denote an edge set of interference links, and a mapping function maps paired nodes to their d - dimensional features , while maps edges to d - dimensional features. In the context of a wireless network graph, node features involve the local channel coefficients of D2D links, including the measurement attributes of the th DG , and . Further define the node feature matrix as , where . Edge features are defined as the channel gains of interference links, which are defined by the set of channel coefficient matrices , and the corresponding adjacent feature element is defined as

[0197]

[0198] Step 1504: Input the node features and edge features into the graph neural network model, use the Lagrangian dual function as the loss function, and combine primal - dual learning to train the graph neural network model. After the training is completed, obtain the output of the graph neural network model to get the optimal voltage control and communication power allocation strategy.

[0199] The uplink information exchange in the wireless network control system structure considered in the embodiments of this application is embedded in fading and interference patterns, which can be incorporated into policy parameterization and aggregated through the message - passing phase, and is designed as the convolution kernel of the message - passing graph neural network. In the message - passing phase, each node generates a message based on its local features and passes the generated message to different neighbors through the edge features of the neighbors. In the aggregation phase, the node aggregates messages from all neighbors to generate aggregated information. Next, in the update phase, the node combines the aggregated information from neighbors and its local features to update its embedded representation. The embedded representation obtained through one update step represents the expression of its local features and neighbor features.

[0200] For each node , let denote the initial feature vector corresponding to node , where represents the number of initial features of each node. These feature vectors are passed through the edges of the graph for multi - layer message passing and aggregation. Let denote the number of layers. For each layer , Represents a node At the layer's feature vector, Indicates the number of features for each node in each layer. Each such feature vector is based on the features of the nodes in the previous layer And its neighbors and self-loops are aggregated to obtain,

[0201]

[0202] Among them, Indicates the potential non-linear aggregation function parameterized by the policy parameter Parameterized, Represents a node The set of neighbors of, defined as , Represents the set of nodes, Represents a node Of the adjacent nodes, Represents the set of edges, Then represents the node And There is an adjacency relationship between. After Layers, each node will have a final feature vector , which is called the embedding representation of the node. Therefore, The output of the layer graph neural network can be expressed as:

[0203]

[0204] Among them , Represents the communication domain power allocation decision, Represents the control domain reactive power ratio decision.

[0205] For each pair of transmitter and receiver, the transmit power level is:

[0206]

[0207] Among them, Represents the Sigmoid function , Represents a parameter vector that maps the average node embedding to a scalar and then converts it to its allocated transmit power.

[0208] For photovoltaic (PV) equipped with a continuous action set, Represents the generated reactive power ratio,

[0209]

[0210] Among them, Represents the designed stacked ReLU function, is a parameter vector used to convert node embeddings into a probability distribution of reactive power ratios.

[0211] By using the hierarchical optimization decomposition method, the dual problem can be decomposed into a two-level hierarchy. Evaluating the dual function is transformed into a Markov decision process (MDP) problem, where the objective is defined by the Lagrangian function equation. Therefore, the inner maximization can be solved using the SPGNN method. The outer optimization problem is designed to minimize the dual problem with respect to That is, the inner layer updates the parameters of the graph neural network, and the outer layer updates the parameters of the Lagrange multipliers, with the policy parameters of the constraint set is

[0212]

[0213] The distributed communication resource allocation and voltage collaborative regulation method provided by the embodiments of the present application parameterizes the constraint conditions in the cross-domain optimization problem to establish the constraint conditions in the design of the graph neural network model, forming a parameterized joint optimization problem; by simultaneously considering the Lagrangian dual function of the optimization objective and the constraint conditions to express the parameterized joint optimization problem, the wireless network control system is modeled as a directed communication graph, the node features and edge features of the directed communication graph are determined, and by inputting the node features and edge features into the graph neural network model, using the Lagrangian dual function as the loss function, and combining primal-dual learning, the graph neural network model is trained. After the training is completed, the output of the graph neural network model is obtained to obtain the optimal voltage control and communication power allocation strategy, which can provide a remote control design information interaction transmission scheme for high new energy access ratios in the distribution network. Considering the requirements of the control system and the limitations of spectrum and energy resources, the optimal voltage control strategy and communication resource allocation strategy are obtained, realizing the collaborative control of distributed communication resource allocation and voltage, and improving the information exchange ability and control performance of the new power system.

[0214] The distributed communication resource allocation and voltage collaborative regulation method provided by the embodiments of the present application can be executed by a distributed communication resource allocation and voltage collaborative regulation device. In the embodiments of the present application, taking the distributed communication resource allocation and voltage collaborative regulation device executing the distributed communication resource allocation and voltage collaborative regulation method as an example, the distributed communication resource allocation and voltage collaborative regulation device provided by the embodiments of the present application is described.

[0215] The embodiments of the present application also provide a distributed communication resource allocation and voltage collaborative regulation device, such as Figure 4As shown, the distributed communication resource allocation and voltage coordinated control device 400 includes: a system construction unit 410, a model determination unit 420, a cross-domain dependency determination unit 430, an optimization problem determination unit 440 and a solution unit 450, wherein:

[0216] A system construction unit 410 is used to construct a wireless network control system based on a 5G local area network for a new power system;

[0217] A model determination unit 420, configured to determine a wireless transmission model and a dynamic model of voltage control of the wireless network control system;

[0218] A cross-domain dependency relationship determination unit 430, configured to determine a cross-domain dependency coupling relationship between wireless communication and control system performance of the wireless network control system based on the wireless transmission model and the dynamic model of voltage control;

[0219] An optimization problem determining unit 440, configured to determine a cross-domain optimization problem based on the cross-domain dependency coupling relationship;

[0220] The solving unit 450 is used to solve the cross-domain optimization problem by using the primal-dual graph neural network algorithm to obtain the optimal voltage control and communication power allocation strategy.

[0221] In some embodiments, the system construction unit 410 is used to:

[0222] The distributed controller is connected to multiple distribution network nodes of the new power system through the 5G local area network to realize distributed communication and control, and dynamically adjust the voltage control and communication power allocation strategy in each control cycle to cope with voltage fluctuations and load changes in the distribution network.

[0223] In some embodiments, the connecting of the distributed controller to multiple distribution network nodes of the novel power system through a 5G local area network includes:

[0224] Determining a plurality of control areas according to a distribution network topology structure corresponding to the novel power system;

[0225] Wireless bearer is realized between different control areas through network nodes;

[0226] In each control area, the network node covers at least one distributed energy node, and the distributed energy node includes: a new energy generation node or a common distributed generation node. The common distributed generation node serves as the distribution network node and is sensed by the associated new energy generation node. The common distributed generation node and the associated new energy generation node communicate through an end-to-end link in the control area to which they belong;

[0227] The status information of the ordinary distributed generation nodes is transmitted to the new energy generation nodes through the end-to-end link, forming the upstream status perception of the control loop. The new energy generation nodes, as the distributed controllers, aggregate the upstream status information through the end-to-end link, determine the reactive power injection ratio value, and execute the downstream voltage control command;

[0228] The network node distributes power for the end-to-end link within the control area covered by the network node and transmits the power distribution instruction to the new energy generation node.

[0229] In some embodiments, the m-th new energy generation node associated with the adjacent ordinary distributed generation nodes is set as the m-th distributed controller of the wireless network control system, forming the m-th control subsystem. Multiple such control subsystems form a closed feedback loop through the 5G local area network;

[0230] Wherein, the status of the m-th new energy generation node at a certain moment depends on its own measurement value and the measurement values obtained from its neighbors.

[0231] In some embodiments, the cross-domain dependency determination unit 430 is configured to:

[0232] Based on the Lyapunov stability theory and the wireless transmission model, perform stability and rapidity analysis on the dynamic model of the voltage control, and determine the exponential stability constraint condition and the convergence rate constraint condition under the imperfect wireless link condition.

[0233] In some embodiments, the optimization problem determination unit 440 is configured to:

[0234] Determine the objective function, which includes: a first objective function for converging the corresponding voltage to the desired range within a preset time period, and a second objective function for introducing penalties on voltage deviation and reactive power injection;

[0235] According to the dynamic model rules, transient response requirements, the exponential stability constraint condition and the convergence rate constraint condition under the imperfect wireless link condition, the transmit power constraint condition, and the packet loss rate constraint condition, determine the objective constraint conditions;

[0236] According to the objective function and the objective constraint conditions, determine the cross-domain optimization problem.

[0237] In some embodiments, the solving unit 450 is configured to:

[0238] Parameterize the constraint conditions in the cross-domain optimization problem to establish the constraint conditions in the design of the graph neural network model, forming a parameterized joint optimization problem;

[0239] Express the parameterized joint optimization problem through the Lagrangian dual function that simultaneously considers the optimization objective and constraints;

[0240] Model the wireless network control system as a directed communication graph, and determine the node characteristics and edge characteristics of the directed communication graph;

[0241] Input the node characteristics and edge characteristics into the graph neural network model, use the Lagrangian dual function as the loss function, and combine primal-dual learning to train the graph neural network model. After the training is completed, obtain the output of the graph neural network model to obtain the optimal voltage control and communication power allocation strategy.

[0242] The distributed communication resource allocation and voltage cooperative regulation device in the embodiments of the present application can be an electronic device or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a processor, a controller, a computing platform, a cloud platform, a server, etc. or a terminal, or other devices, and the embodiments of the present application do not make specific limitations.

[0243] The distributed communication resource allocation and voltage cooperative regulation device in the embodiments of the present application can be a device with an operating system. The operating system can be the Microsoft (Windows) operating system, the Android operating system, the IOS operating system, or other possible operating systems, and the embodiments of the present application do not make specific limitations.

[0244] The distributed communication resource allocation and voltage cooperative regulation device provided by the embodiments of the present application can implement Figures 1 to 3 each process implemented by the method embodiments, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0245] In some embodiments, as Figure 5 shown, the embodiments of the present application also provide an electronic device 500, including a processor 501, a memory 502, and a computer program stored on the memory 502 and executable on the processor 501. When the program is executed by the processor 501, it implements each process of the above-mentioned distributed communication resource allocation and voltage cooperative regulation method embodiments, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0246] It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0247] The embodiments of the present application also provide a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements each process of the above-mentioned embodiments of the distributed communication resource allocation and voltage collaborative regulation method, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0248] Wherein, the processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media such as computer read-only memory ROM, random access memory RAM, magnetic disk or optical disc, etc.

[0249] The embodiments of the present application also provide a computer program product, including a computer program. When the computer program is executed by a processor, it implements the above-mentioned distributed communication resource allocation and voltage collaborative regulation method.

[0250] Wherein, the processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media such as computer read-only memory ROM, random access memory RAM, magnetic disk or optical disc, etc.

[0251] The embodiments of the present application further provide a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement each process of the above-mentioned embodiments of the distributed communication resource allocation and voltage collaborative regulation method, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0252] It should be understood that the chip mentioned in the embodiments of the present application can also be referred to as a system-on-chip, system chip, chip system or system-on-chip, etc.

[0253] It should be noted that in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including that element. In addition, it should be pointed out that the methods and devices in the embodiments of the present application are not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0254] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases, the former is a better implementation. Based on such an understanding, the technical solution of the present application, in essence or the part that contributes to the related art, can be embodied in the form of a computer software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to enable a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present application.

[0255] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms without departing from the purpose of the present application and the scope protected by the claims, and all of them belong to the protection scope of the present application.

[0256] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0257] Although the embodiments of the present application have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principle and purpose of the present application. The scope of the present application is defined by the claims and their equivalents.

Claims

1. A distributed communication resource allocation and voltage coordinated control method, characterized in that: include: For new power systems, build a wireless network control system based on 5G local area network; Determining a wireless transmission model and a dynamic model of voltage control of the wireless network control system; Based on the wireless transmission model and the dynamic model of voltage control, determining the cross-domain dependent coupling relationship between wireless communication and control system performance of the wireless network control system; Based on the cross-domain dependency coupling relationship, determining a cross-domain optimization problem; The cross-domain optimization problem is solved by using the primal-dual graph neural network algorithm to obtain the optimal voltage control and communication power allocation strategy; The wireless network control system based on 5G local area network is constructed for the new power system, including: Connecting the distributed controller to multiple distribution network nodes of the novel power system through the 5G local area network to realize distributed communication and control, and dynamically adjusting the voltage control and communication power allocation strategy in each control cycle to cope with voltage fluctuations and load changes in the distribution network; Among them, the status information of the distribution network node is transmitted to the distributed controller to form an uplink status perception of the control loop. The distributed controller determines the reactive power injection ratio value by aggregating the uplink status information, executes the downlink voltage control command, and receives the power allocation instruction.

2. The distributed communication resource allocation and voltage coordinated control method according to claim 1, characterized in that: The method of connecting the distributed controller to multiple distribution network nodes of the novel power system through the 5G local area network includes: Determining a plurality of control areas according to a distribution network topology structure corresponding to the novel power system; Wireless bearer is realized between different control areas through network nodes; In each control area, the network node covers at least one distributed energy node, and the distributed energy node includes: a new energy generation node or a common distributed generation node. The common distributed generation node serves as the distribution network node and is sensed by the associated new energy generation node. The common distributed generation node and the associated new energy generation node communicate through an end-to-end link in the control area to which they belong; The state information of the common distributed power generation node is transmitted to the new energy power generation node through the end-to-end link to form an uplink state perception of the control loop. The new energy power generation node acts as the distributed controller to aggregate the uplink state information through the end-to-end link, determine the reactive power injection ratio value, and execute the downlink voltage control command; The network node performs power allocation for end-to-end links within a control area covered by the network node, and transmits a power allocation instruction to the new energy power generation node.

3. The distributed communication resource allocation and voltage coordinated control method according to claim 2, characterized in that: The mth new energy generation node associated with the adjacent common distributed generation node is set as the mth distributed controller of the wireless network control system to form an mth control subsystem, and a plurality of the control subsystems form a closed feedback loop through the 5G local area network; The state of the m-th new energy generation node at a certain moment depends on its own measurement value and the measurement value obtained from its neighbors.

4. The distributed communication resource allocation and voltage coordinated control method according to claim 1, characterized in that: The determining of the cross-domain dependent coupling relationship between the wireless communication and the control system performance of the wireless network control system based on the wireless transmission model and the dynamic model of voltage control includes: Based on Lyapunov stability theory and the wireless transmission model, the stability and rapidity of the voltage-controlled dynamic model are analyzed to determine the exponential stability constraint and convergence rate constraint under the condition of an imperfect wireless link.

5. The distributed communication resource allocation and voltage coordinated control method according to claim 4 is characterized in that: Determining the cross-domain optimization problem based on the cross-domain dependency coupling relationship includes: Determining an objective function, the objective function comprising: a first objective function for causing the corresponding voltage to converge to a desired range within a preset time period, and a second objective function for introducing penalties for voltage deviation and reactive power injection; Determining target constraints according to dynamic model rules, transient response requirements, exponential stability constraints and convergence rate constraints under the imperfect wireless link conditions, transmission power constraints, and packet loss rate constraints; The cross-domain optimization problem is determined according to the objective function and the objective constraint condition.

6. The distributed communication resource allocation and voltage coordinated control method according to claim 1, characterized in that: The cross-domain optimization problem is solved by using the primal-dual graph neural network algorithm to obtain the optimal voltage control and communication power allocation strategy, including: Parameterizing the constraints in the cross-domain optimization problem to establish the constraints in the design of the graph neural network model to form a parameterized joint optimization problem; The parameterized joint optimization problem is expressed by a Lagrangian dual function that considers both the optimization objective and the constraints; Modeling the wireless network control system as a directed communication graph, and determining node characteristics and edge characteristics of the directed communication graph; The node features and edge features are input into the graph neural network model, and the Lagrangian dual function is used as the loss function. The graph neural network model is trained in combination with the original dual learning. After the training, the output of the graph neural network model is obtained to obtain the optimal voltage control and communication power allocation strategy.

7. A distributed communication resource allocation and voltage coordinated control device, characterized in that: include: System construction unit, used to build a wireless network control system based on 5G local area network for new power systems; A model determination unit, used to determine a wireless transmission model and a dynamic model of voltage control of the wireless network control system; A cross-domain dependency relationship determination unit, used to determine the cross-domain dependency coupling relationship between the wireless communication of the wireless network control system and the control system performance based on the wireless transmission model and the dynamic model of voltage control; An optimization problem determination unit, configured to determine a cross-domain optimization problem based on the cross-domain dependency coupling relationship; A solving unit, used for solving the cross-domain optimization problem by using a primal-dual graph neural network algorithm to obtain an optimal voltage control and communication power allocation strategy; The wireless network control system based on 5G local area network is constructed for the new power system, including: Connecting the distributed controller to multiple distribution network nodes of the novel power system through the 5G local area network to realize distributed communication and control, and dynamically adjusting the voltage control and communication power allocation strategy in each control cycle to cope with voltage fluctuations and load changes in the distribution network; The status information of the distribution network node is transmitted to the distributed controller to form an uplink status perception of the control loop. The distributed controller determines the reactive power injection ratio value by aggregating the uplink status information, executes the downlink voltage control command, and receives the power allocation instruction.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the distributed communication resource allocation and voltage coordinated control method as described in any one of claims 1-6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the distributed communication resource allocation and voltage coordinated control method as described in any one of claims 1 to 6 is implemented.

Images