Virtual power plant adjustable potential comprehensive evaluation method, system and equipment giving consideration to communication and action time delay

Abstract

The invention discloses a virtual power plant adjustable potential comprehensive evaluation method, system and equipment giving consideration to communication and action time delay. According to the method, a communication system time delay model, an equipment action time model and a communication reliability model are constructed by constructing a cloud edge collaborative network system framework for regulating and controlling a virtual power plant. And then, establishing a virtual power plant maximum adjustable potential evaluation model based on timing delay step length search by using a k-shortest path algorithm. And finally, providing a method based on a near-end strategy optimization algorithm to solve the optimization model. According to the method, comprehensive evaluation of the adjustable potential of the virtual power plant considering the influence of the communication system and the resource action time is accurately described, a decision basis is provided for the virtual power plant to participate in primary and secondary frequency modulation and peak regulation of the power system, and frequency stability and power and electric quantity balance of the power system can be supported.

Description

Technical Field

[0001] The present invention relates to the field of comprehensive evaluation of the adjustable potential of flexible loads, and in particular to a method, system and device for comprehensive evaluation of the adjustable potential of a virtual power plant taking into account communication and action delays. Background Art

[0002] As an emerging technology and market player, virtual power plants are considered to be an effective means to alleviate the peak-to-valley difference of power grid load and maintain the balance of power grid supply and demand in the future. The quantitative relationship between the response speed and adjustable capacity of virtual power plants can reflect their ability to respond to system regulation needs at different time scales, which is mainly affected by communication time and resource regulation action performance. Therefore, the comprehensive evaluation of the adjustable potential of virtual power plants considering the influence of communication system and resource action time can provide a decision-making basis for their participation in primary and secondary frequency regulation and peak regulation of power systems, which is beneficial to their support for frequency stability and power balance of power systems.

[0003] Many scholars have conducted research in the field of comprehensive evaluation of the adjustable potential of flexible loads, but existing research cannot meet the needs of actual grid dispatching. Actual grid dispatching also requires the evaluation of comprehensive regulation capabilities such as response speed and duration. In addition, the traditional centralized architecture used will bring a series of disadvantages at the communication transmission level. At the same time, existing research has not taken into account the problem of deviation in response capacity caused by factors such as delay, packet loss, and bit error during information transmission. Summary of the invention

[0004] Purpose of the invention: The present invention aims to provide a method for comprehensively evaluating the adjustable potential of flexible loads of virtual power plants taking into account both response capacity and response delay from the perspective of physical information systems, taking into account various influencing factors such as equipment action response delay, communication reliability, packet loss and bit errors during communication; another object of the present invention is to provide a system and device for comprehensively evaluating the adjustable potential of virtual power plants taking into account communication and action delays.

[0005] Technical solution: The comprehensive evaluation method for the adjustable potential of a virtual power plant taking into account communication and action delays of the present invention comprises the following steps:

[0006] According to the virtual power plant cloud-edge collaborative network system framework, the communication system delay model, equipment action time model and communication reliability model are constructed respectively;

[0007] According to the virtual power plant cloud-edge collaborative network system framework, communication system delay model, equipment action time model and communication reliability model, the KSP algorithm is used to obtain the optimal path and backup path between the source node m and the destination node n, and determine the total transmission delay t between the source node m and the destination node n during data transmission. d , construct a maximum adjustable potential evaluation model for virtual power plants based on timed delay step search;

[0008] The PPO algorithm is used to optimize the maximum adjustable potential evaluation model of the virtual power plant, and an online evaluation of the adjustable potential of the virtual power plant is obtained.

[0009] Furthermore, the virtual power plant cloud-edge collaborative network system framework includes a cloud layer, an edge layer and a response terminal; the cloud layer is used to centrally integrate the computing resources of many servers and is responsible for processing global data services, including comprehensive evaluation, prediction of the aggregate adjustable capacity of virtual power plants, decomposition of scheduling instructions to each edge node, and overall communication optimization control of the cloud-edge-end system; the edge layer is used to calculate the aggregate load resource model at the edge nodes, and the edge nodes of the edge layer are set close to the user side.

[0010] Furthermore, the communication system delay model is as follows:

[0011] Wired transmission delay between the central cloud and edge nodes for

[0012]

[0013] Among them, x i As a binary variable, x i =1 means that there is data packet transmission between cloud node and edge node i; otherwise x i =0;L i represents the size of the downlink data packet of the i-th node; r i is the wired transmission rate between the central cloud and the edge nodes; b i represents the channel bandwidth between cloud nodes, P i represents the data packet transmission power, G i represents the channel gain, and G i =d i -α , d i represents the distance between the i-th node and the central cloud node, -α is the signal attenuation factor, and s represents the transmission noise;

[0014] Wired propagation delay between cloud nodes and edge nodes for

[0015]

[0016] Where v is the propagation speed of the optical signal in the optical fiber;

[0017] Downlink data delay between cloud edge and edge node t i for

[0018]

[0019] Among them, Nh Indicates the number of routing hops, t f Indicates the processing delay of a single device;

[0020] The jth edge node calculates the delay for

[0021]

[0022] Among them, f j is the computing power of the jth edge node, e j The number of CPU cycles required to calculate the task;

[0023] Wireless download delay t p for

[0024]

[0025] in, represents user p in the jth region node j The queue position where it is located, T is the smallest indivisible period in wireless transmission;

[0026] Wireless delay t p Using Pareto distribution to describe, the probability distribution function P is

[0027]

[0028] Among them, τ 0 represents the minimum delay between nodes, γ is a parameter related to the load rate, and the cloud server calculates the delay is a constant, t p Wireless delay is generated in the process of transmitting from the base station to the demand response terminal.

[0029] Furthermore, the communication reliability model is as follows:

[0030] The reliability S between nodes m and n mn for

[0031]

[0032] Among them, N mn represents the number of links between m and n; S(e l ) represents link e l The reliability of the link. Information component failure, transmission packet loss, and bit error will affect the reliability of the link. mn The upper limit of the route hop count is set to

[0033] Furthermore, the KSP algorithm is used to obtain the optimal path and backup path between the source node m and the destination node n. The deviation link weight of the transmission path between m and n is as follows:

[0034]

[0035] Among them, W mn is the link weight between nodes m and n, t mn is the delay between nodes m and n, c l represents the packet loss coefficient, Indicates rounding up, t s is the delay caused by packet loss and bit error, which is a part of the total delay and is set to a constant. d is a random number generated between 0 and 1.

[0036] Furthermore, the total transmission delay t between the source node m and the destination node n during data transmission is d for

[0037]

[0038] Among them, t ij The dispatching instruction sent to the edge node generates wired delay through the wireless base station, t a is the action time of the device, t s is the delay caused by packet loss and bit errors.

[0039] Furthermore, the maximum adjustable potential evaluation model of the virtual power plant based on the timed delay step search is:

[0040]

[0041] Among them, F is the total adjustable capacity of the central cloud virtual power plant, Indicates that the pth wireless base station is within the coverage area of ​​the jth wireless base station. j The maximum adjustable capacity of a demand response terminal, t allow is the maximum allowable response delay, x ij It is the path selection result between edge nodes and BS nodes in the load scheduling communication link.

[0042] The virtual power plant adjustable potential comprehensive evaluation system taking into account communication and action delays of the present invention comprises:

[0043] The model building module is used to build a communication system delay model, a device action time model and a communication reliability model according to the virtual power plant cloud-edge collaborative network system framework; it is used to use the KSP algorithm to obtain the optimal path and backup path between the source node m and the destination node n according to the virtual power plant cloud-edge collaborative network system framework, the communication system delay model, the device action time model and the communication reliability model, and determine the total transmission delay t between the source node m and the destination node n during the data transmission process. d , construct a maximum adjustable potential evaluation model for virtual power plants based on timed delay step search;

[0044] The strategy optimization evaluation module is used to determine the optimal strategy for the maximum adjustable potential evaluation model of the virtual power plant through the PPO algorithm, and obtain an online evaluation of the adjustable potential of the virtual power plant.

[0045] The computer device of the present invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0046] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program implements the steps of the above method when executed by a processor.

[0047] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: 1. The present invention adopts a cloud-edge collaborative network system framework to solve the problems of large data transmission delay, difficulty in storing and utilizing massive data, and easy transmission failure and limited communication capacity in the centralized architecture. The edge nodes provide computing resources at the edge of the network, which is closer to terminal devices and users, helping to reduce the computing pressure of cloud nodes, reduce communication network delay, and enhance network information security; 2. The present invention can consider the impact of cloud-edge architecture communication and equipment action time on the adjustable capacity response delay, quantify the regulation performance of short-time scale virtual power plants, and solve the problem that current research cannot meet the actual scheduling needs of the power grid; 3. The present invention takes into account the problem of deviation in response capacity during the scheduling process caused by delays, packet loss, bit errors and other problems during information transmission into the field of comprehensive evaluation of flexible load adjustable potential, so that the evaluation results are more reliable; 4. The present invention uses the k shortest path KSP algorithm to iterate the optimal path and backup path from the source node to the destination node according to the delay parameters of each edge of the network. It provides decision makers with more flexibility and optimization space, and can more effectively deal with uncertain factors such as network congestion and failures when facing complex and changeable network environments; 5. The present invention adopts a virtual power plant adjustable potential evaluation method based on a proximal strategy optimization algorithm to solve the maximum adjustable potential evaluation model of the virtual power plant, and realizes a rapid online evaluation of the adjustable potential of the virtual power plant that takes into account both response capacity and response delay, significantly improves strategy performance and sample efficiency, and effectively avoids drastic performance fluctuations during training. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flow chart of the present invention;

[0049] Figure 2 This is a diagram of the cloud-edge collaborative data interaction architecture;

[0050] Figure 3 This is a diagram of the cloud-edge-end data interaction framework;

[0051] Figure 4 It is a flow chart for adjustable potential assessment and scheduling;

[0052] Figure 5 It is the temperature control load response mode diagram;

[0053] Figure 6 Approximate action time model diagrams for various devices;

[0054] Figure 7 Flowchart for assessing the scalability potential of virtual power plants;

[0055] Figure 8 It is a schematic diagram of the IEEE33 bus system including the communication network;

[0056] Fig. 9 It is the load typical day adjustable potential curve;

[0057] Fig.10 The reward curve diagram for each algorithm;

[0058] Fig.11 This is the result diagram of the daily adjustable potential assessment of the virtual power plant;

[0059] Fig.12 It is the delay proportion diagram of each part of BS node;

[0060] Fig.13 This is the evaluation result diagram of load adjustable potential of each mode;

[0061] Fig.14 Adjustable potential deviation map caused by packet loss;

[0062] Fig.15 This is a graph showing the wireless transmission delay deviation caused by the load rate. DETAILED DESCRIPTION

[0063] The present invention will be further described below in conjunction with the accompanying drawings.

[0064] The method for comprehensively evaluating the adjustable potential of a virtual power plant taking into account communication and action delays of the present invention comprises the following steps:

[0065] (1) Build a cloud-edge collaborative network system framework for regulating virtual power plants;

[0066] In the Outline for the Construction of Ubiquitous Power Internet of Things, State Grid Corporation of China divides the construction of ubiquitous power Internet of Things into four levels: application layer, platform layer, network layer and perception layer. This principle can represent the information-physical architecture of virtual power plants. Based on the above information-physical architecture, Figure 2As shown, the present invention proposes a virtual power plant cloud-edge collaborative architecture to evaluate and predict the adjustable capacity of the massive flexible loads participating in the virtual power plant and realize the optimal scheduling of internal resources. Different from the centralized architecture, the edge node is placed close to the user side. The central cloud relies on the virtual power plant cloud to centrally integrate the computing power resources of many servers and is responsible for processing the global high-complexity data services, including comprehensive evaluation and prediction of the aggregated adjustable capacity of the virtual power plant, decomposition of scheduling instructions to each edge node, and overall communication optimization control of the cloud-edge-end system. Technical service providers and load aggregators are located at the edge nodes, and the edge servers they use have the characteristics of low energy consumption, low latency, and fast reading and writing speed, which are suitable for computing terminal-oriented tasks with high reliability and low latency requirements. The edge nodes are connected to the central cloud and the distribution substation through the power backbone network, and the networking mode combining optical fiber communication, dense wavelength division multiplexing technology and synchronous digital system is mostly adopted. The distribution substation connects to the response terminal through the power access network, usually using LTE, 4 / 5G, Wi-Fi, Ethernet and other communication methods. This architecture makes full use of cloud and edge computing resources, combines the three-layer control requirements of the virtual power plant cloud, edge and user terminals, and meets the needs of control latency, reliability, accuracy and computing efficiency.

[0067] like Figure 3 The cloud-edge-end data interaction framework applied to virtual power plants is shown in the figure. Data from each layer is collected in the cloud center. By calculating and analyzing the collected data, the cloud layer handles the scheduling control and external cooperation services of the virtual power plant from the system level. In the scheduling process at all levels, the model center of the cloud layer is responsible for storing and calculating the global, highly complex model that aggregates all edge layers, so as to ensure the global optimization scheduling within the virtual power plant. The model center of the edge layer is responsible for calculating the aggregated load resource model near the edge node to share the computing pressure of the cloud center, reduce the large amount of delay caused by the direct interaction between the cloud layer and the terminal data in centralized scheduling, and ensure the data privacy security of virtual power plant service users.

[0068] Cloud layer. The virtual power plant cloud server has rich storage and computing resources and can interact with platforms such as the power dispatch center, power grid cloud platform, and power grid data center. In the comprehensive evaluation process of the aggregated adjustable capacity of the virtual power plant, the cloud server is mainly responsible for using the communication system optimization method to plan the information transmission path for the network data within the system to optimize the latency and reliability of data transmission in the information system, and evaluate the aggregated adjustable capacity of the virtual power plant that takes into account the communication and action time based on the uploaded aggregate terminal cluster data. In the process of adjustable capacity prediction, the edge evaluation data is transmitted to the edge prediction model, and the cloud evaluation data is input into the aggregated adjustable capacity prediction model at the same time, and the cloud-edge mechanism is used to realize refined adjustable capacity prediction. Based on external data such as electricity price data and unit data, the cloud platform performs virtual power plant participation in the economic dispatch calculation of the power system to ensure economic dispatch based on the balance of supply and demand. Then, based on the obtained day-ahead and intraday dispatch plans, the optimal dispatch of the adjustable resources within the virtual power plant is guided.

[0069] Edge layer. The edge layer plays the role of connecting the terminal and the cloud layer in cloud-edge collaboration. The edge node has both the uplink communication capability with the virtual power plant cloud platform and the downlink communication capability with the distribution substation, including edge controllers, edge servers, and edge gateways. Each edge node is not only responsible for the data processing and operation management of the distribution substation area where the regional data is uploaded, but also has the function of assisting the analysis of large models deployed on the cloud server in the evaluation and prediction of the adjustable capacity of edge aggregation. In the evaluation of the adjustable capacity of virtual power plants, edge nodes are responsible for collecting uploaded user operation data, evaluating the adjustable capacity of users in the region based on the differentiated evaluation model configured on the edge side, and uploading cluster data, which is further evaluated by the cloud platform. In addition, edge nodes can also be configured with small user-oriented applications, such as short-term load baseline forecasting, demand response plan filling, etc., to help users fine-tune future electricity consumption behaviors.

[0070] Terminal. The response terminal equipment is deployed on the user side, responsible for real-time collection of power data and equipment operation data and uploading them to the distribution substation. The user terminal and the distribution substation are in a many-to-one correspondence. The terminal is mainly intelligent-oriented, controls multiple deployed intelligent systems, and comprehensively regulates the operation of various intelligent electrical equipment according to the downloaded terminal control instructions.

[0071] The process of evaluating and dispatching the adjustable potential of a virtual power plant based on cloud-edge collaboration is as follows:

[0072] 1) The automatic demand response contract is signed by the user, and the user data is sent to the base station through the regional network at regular intervals.

[0073] 2) The edge node EN collects data on customer loads in the coverage area, predicts the adjustable capacity and response time of the adjustable load based on real-time and historical data, and uploads the results to the cloud node.

[0074] 3) The central cloud plans the information transmission path based on the network topology and evaluates the adjustable potential of the area including reliability and latency metrics based on the uploaded data.

[0075] 4) When an imbalance in power grid supply and demand is detected, the cloud optimization results are calculated and sent to the EN. After edge optimization, the scheduling instructions are transmitted to the terminal, which executes and feeds back the results.

[0076] like Figure 4 The figure shows the adjustable potential evaluation and scheduling process, and the two-stage scheduling process of the negative virtual power plant under the cloud-edge collaborative architecture, which is divided into three stages as a whole. The first stage is the data upload stage. Each terminal node collects the real-time operation data of the load within the node range, and the regional node collects and processes the regional operation data; the second stage is the data processing stage. After the cloud node receives the scheduling instruction issued by the power dispatching center, it sends a request to the regional node. The regional node sends the regional data to the selected edge node according to the preset path with the smallest upload delay. The edge node evaluates the real-time adjustable capacity of each region and calculates the active power and reactive power of each region. The data calculation results are sent to the corresponding regional node and cloud node at the same time; the third stage is the optimization and instruction issuance stage, which is the optimization of the response instruction to the terminal process in the actual scheduling process. The optimization process is divided into two-stage optimization of cloud and edge. The cloud optimization includes two parts. The first part is the response power optimization of the regional node to obtain the total demand response scheduling power in each region; the second part is the optimization of the information transmission path to select the terminal node corresponding to the edge computing and plan the transmission path. The node selection information is sent to the regional node, and the regional node uploads the adjustable capacity evaluation results and power data to the edge node according to the planned route to prepare for edge optimization. The object of edge optimization is the response power of the terminal node. The response instruction is sent to the terminal node and executed by the terminal.

[0077] The present invention establishes a cloud-edge collaborative system in which the network topology is represented by a weighted undirected graph G={V,E}. is the set of nodes in the network, N v Indicates the number of nodes, is the link set of the network, N e represents the number of links. The network nodes include the central cloud server, edge server, distribution substation and router. The set of I server nodes is recorded as When i=0, it indicates the cloud server located in the regional main substation. Assume that J distribution substations distributed in various residential areas and commercial buildings have deployed wireless base stations (BS), and the set is represented as The wireless coverage area of ​​the jth BS contains P j user demand response terminals, whose set is expressed as Data is transmitted between the base station and the server, as well as between server nodes, via an optical fiber network, and data exchange is completed through routers.

[0078] (2) Based on the cloud-edge collaborative network system framework, models of communication delay, device action time, and communication reliability were constructed respectively;

[0079] The communication system delay model is as follows:

[0080] After receiving the scheduling task, the cloud server first sends a control instruction to the edge node to generate the wired downlink data delay t i In the distributed optimization process under the cloud-edge information system architecture, the cloud server optimizes the communication path and the total demand response in each BS coverage area, and the edge node optimizes the responsiveness of each user demand response terminal. The delay generated in this process includes the computing delay of the cloud server. and edge node computing latency The edge node sends a scheduling instruction, which generates a wired delay t through the BS ij , and finally the wireless delay t is generated during the transmission from the base station to the demand response terminal p Since the user response return data is small in size, it can be ignored. The device action time is expressed as t a .

[0081] The end-to-end delay in the power system includes inherent delay, transmission delay, propagation delay and queuing delay. Since the inherent delay accounts for a very small part of the total delay, it can be ignored. According to Shannon's theorem, the wired transmission rate r between the central cloud and the edge node is i for:

[0082]

[0083] Where: x i As a 01 variable, x i =1 means that there is data packet transmission between cloud node and edge node i, otherwise x i =0,L i represents the size of the downlink data packet of the i-th node, b i represents the channel bandwidth between cloud nodes, P i represents the data packet transmission power, G i represents the channel gain, and G i =d i -α , d i represents the distance between the ith node and the central cloud node, -α is the signal attenuation factor, and s represents the transmission noise.

[0084] The propagation delay is related to the end-to-end distance. The wired propagation delay between the cloud node and the edge node EN is It can be expressed as:

[0085]

[0086] Where: v is the propagation speed of the optical signal in the optical fiber, which is generally 2 / 3c. Downlink data delay between cloud edge and EN t i It can be expressed as follows:

[0087]

[0088] Where: N h Indicates the number of routing hops, t f Indicates the processing latency of a single device.

[0089] For the wired delay t between EN and BS ij , its calculation principle is consistent with the calculation method of the finite delay between the cloud node and EN, which is the same as formula (3).

[0090] The cloud server allocates the optimization tasks to EN according to the edge optimization requirements after optimizing in the cloud according to the dispatching instructions issued by the power dispatching center. The computing power of the jth edge node is defined as f j , the number of CPU cycles required for the calculation task is e j , its computational delay can be expressed as:

[0091]

[0092] The delay caused by wireless transmission includes the upload delay of the data upload process and the download delay caused by the instruction download process. The wireless upload delay can be calculated using the M / M / 1 queuing model. However, in this paper, since the process of uploading user data to the regional node is carried out in advance, the wireless upload delay is not considered. During the download process, the regional node issues scheduling instructions based on the load adjustable capacity. The terminal with a larger adjustable capacity has a higher ranking in the queue. Since the data packets transmitted during the download phase are small, the wireless download delay t p It can be expressed as:

[0093]

[0094] In the formula represents user p in the jth region node j The queue position where T is located is the smallest indivisible period in wireless transmission.

[0095] In addition, the computing resources of general cloud servers are large, so the computing delay of cloud servers can be set is a constant. The delay of the information link between two nodes in the access network will show a probabilistic characteristic. When the service load is certain, the wireless transmission delay can be described by Pareto distribution, and the probability distribution function is expressed as:

[0096]

[0097] Where: τ 0 It represents the minimum delay between nodes, and γ is a parameter related to the load rate.

[0098] The device action time model is as follows:

[0099] The present invention divides user equipment in potential assessment into four types: electric vehicles, temperature-controlled loads, non-industrial interruptible loads and industrial loads.

[0100] 1) Electric vehicles

[0101] The operator can arrange for the electric vehicles parked at the charging station to perform equivalent energy storage and communicate with the BS through the access point. It is assumed that electric vehicles can only be parked in the electric vehicle cluster and the parking status of the electric vehicle remains unchanged during the scheduling. After receiving the scheduling signal, the charging station sets the target electric vehicle charging capacity and broadcasts the data packet to all electric vehicles. The corresponding electric vehicles receive the data packet and respond.

[0102] 2) Temperature control load

[0103] Temperature control loads can provide operating reserves by reducing power consumption. Typical temperature control loads are air conditioning loads and electric water heaters. During normal operation, the temperature control load has two states: operating mode and standby mode, and the two operating powers are alternated. Figure 5 As shown in the figure, taking the cooling mode as an example, when the room temperature is lower than the set maximum allowable temperature, the load operating power is increased from P set1 Down to P set2 , thus providing adjustable capability. When the temperature control load is at t sig After receiving the load reduction control signal, the state of the machine is (a) cooling mode and (b) standby mode. When in cooling mode, its operating power P se t1 immediately changes to standby power P std , due to the hysteresis effect of indoor temperature change, at t re When the room temperature reaches the set temperature, the temperature control load power is converted back to the new operating power P set2 , in this case the conditioning is equivalent to executing the response plan almost immediately. sig When in standby mode, the temperature control load maintains the standby power P std , after the room temperature rises at t re If the response plan is not executed, it will be restarted at t o The cooling mode of the original set temperature is entered at all times. At this time, the equivalent action time of the temperature control load is t ref It can be expressed as:

[0104] t ref =t o -t sig (8)

[0105] 3) Interruptible load

[0106] There are many types of interruptible loads, which are relatively scattered, including lighting equipment, kitchen equipment, washing machines, etc. Such loads are allowed to be conditionally powered off under the incentive mechanism, and loads in the translation operation period are also included in this category. Its control strategy adopts switch control, which has a fast response speed, usually in milliseconds to seconds. However, due to the large differences in types, adjustment accuracy and communication environment between different loads, the adjustable capacity of the aggregation group is a function that increases over time.

[0107] 4) Industrial load

[0108] Industrial loads account for the largest proportion of adjustable capacity in demand response, because their signal receiving and control devices have high precision and the stability of the response process is also the highest. Some industrial loads can respond to the response signal immediately and have the function of continuous adjustment, such as electrolytic aluminum, steel industry and pumping station. The climbing rate after the response is small, and the action time ranges from 30 seconds to several minutes; another part of the industrial load needs to be kept on or off for a long time, including hydraulic presses, forging presses, grinders, etc. This type of load is not suitable for participating in system frequency modulation, so this invention does not consider it.

[0109] In summary, the approximate action time models of various devices are as follows: Figure 6 As shown in the figure, the relationship between the action time and the adjustable capacity is described. The adjustable capacity of the load increases with time after the action time of the load and finally reaches its target adjustable capacity. Since the target adjustable capacity, action time and climbing rate of different types of loads in practice are different, the starting point, steepness and highest point of each curve in the figure are also different. The specific value should be obtained from industry research, and this figure is only a schematic diagram. For the convenience of calculation, the present invention represents the action time of the equipment as t a .

[0110] The communication reliability model is as follows:

[0111] The reliability of a communication system represents the probability of error-free transmission in communication. The reliability S between nodes m and n mn It can be expressed as:

[0112]

[0113] Where: N mn represents the number of links between m and n, S(e l ) represents link e lThe reliability of the link is affected by information element failure, transmission packet loss, and bit errors. Since there is usually more than one communication link from the BS to the cloud node and between cloud nodes, when a node or link fails, the continuity will not be affected by the switching of the backup path, and additional switching delay will be generated. When all information transmission paths between nodes are interrupted, information transmission interruption will occur. When a data packet is lost during transmission, the sender resends the data packet after a period of time due to the influence of the retransmission mechanism. This process will generate a retransmission delay t s In terms of transmission errors, the probability of bit errors in physical information systems is extremely low, generally no more than 10 5 , so the errors in information transmission caused by bit errors can be ignored. In addition, as the number of routing hops increases, the routing reliability will decrease, so an upper limit on the number of routing hops is set. constraints.

[0114] (3) According to the cloud-edge collaborative network system framework and the models of communication delay, equipment action time and communication reliability, a virtual power plant maximum adjustable potential evaluation model based on timing delay step search was established;

[0115] Unlike centralized dispatching, the cloud-edge information system adds edge nodes close to the user side, dispatches user terminals within the management range of the edge nodes, and interacts with cloud servers to reduce the latency caused by uploading and downloading power data. In order to accurately quantify the demand response dispatch latency under the cloud-edge communication system, the power grid demand-side management platform needs to evaluate the load resource dispatch potential under the influence of the communication system in real time, and present the relationship curve between response time and adjustable capacity, to provide support for the precise dispatch of demand response services. When the power grid is unbalanced, operators can dispatch virtual power plants according to precise latency requirements.

[0116] In order to determine the total delay t during data transmission d , the first task is to determine the path for data transmission, and the choice of path depends on the size of the delay of different paths. The present invention uses the KSP algorithm to iterate the optimal path and backup path from the source node to the destination node according to the delay parameters of each edge of the network. The KSP algorithm is a form of path search algorithm, which uses the deviation path method to obtain the shortest path set between two nodes in the cloud-edge collaborative information network, so that the optimal path that meets the needs can be selected. The present invention designs a KSP algorithm for the purpose of minimizing data transmission deviation, taking into account the delay, packet loss and bit error problems in information nodes and transmission links, and uses the KSP algorithm to iterate the optimal path and backup path from the source node to the destination node according to the delay parameters of each edge of the network. At the same time, the delay t caused by packet loss and bit error s It is also considered as part of the transmission deviation. The transmission deviation link weight between nodes m and n is defined as follows: The delay caused by packet loss ts It is also considered as part of the total delay. The link weight between nodes m and n is defined as follows:

[0117]

[0118] Where: c l Indicates the packet loss coefficient, which is determined by the network traffic. mn is the delay between m and n. Indicates rounding up. d is a random number generated between 0 and 1.

[0119] In the KSP calculation process, the network topology matrix is ​​first input. After receiving a request containing source node m and destination node n, the KSP algorithm searches for k shortest paths according to link weights and searches for backup paths after removing the corresponding nodes.

[0120] When the KSP algorithm iterates the optimal path and backup path from the source node to the destination node based on the delay parameters of each edge of the network, the data transmission path is determined accordingly, that is, the path with the lowest node occupancy rate among the shortest paths iterated by the KSP algorithm is preferentially selected as the data transmission path. If there is a transmission failure, it switches to the backup path. When the transmission path is determined, the transmission delay between the two nodes during the data transmission process can be determined, that is, the total transmission delay t between the source node m and the destination node n d It can be expressed as:

[0121]

[0122] In order to evaluate the adjustable capacity and dispatch delay of terminal load in each period, the optimization objective function is the total adjustable capacity F of the central cloud virtual power plant under the allowed delay, and the optimization variable is the path selection result x between the edge node and the BS node in the load dispatch communication link. i and x ij In each evaluation, the maximum allowable response delay t allow The maximum adjustable capacity under the discontinuous delay sequence is obtained by increasing the step length at the same time. The optimization process is shown in the following formula:

[0123]

[0124] Where: F jpj Indicates that the pth BS is within the coverage area of ​​the jth BS. j The maximum adjustable capacity of a demand response terminal. is the upper limit of the number of routing hops, which is a constraint set based on the communication reliability described above.

[0125] (4) According to the cloud-edge collaborative network system framework in step (1), the models of communication delay, equipment action time and communication reliability in step (2), and the maximum adjustable potential evaluation model of the virtual power plant based on the timing delay step search in step (3), a maximum adjustable potential evaluation method based on the proximal policy optimization (PPO) algorithm is constructed to solve the maximum adjustable capacity of the model in step (3);

[0126] Due to the complexity of virtual power plants and communication environments in large-scale links, the present invention proposes an adjustable potential evaluation method based on a proximal policy optimization algorithm on the basis of a fixed-step virtual power plant adjustable potential evaluation model to solve the maximum total adjustable capacity of the virtual power plant, thereby realizing real-time evaluation of the adjustable potential of the virtual power plant. PPO, as a deep reinforcement learning method, can be regarded as a system in which an intelligent agent interacts with the environment. The adjustable potential evaluation problem can be described as an intelligent agent finding a set of actions that maximizes the total adjustable load in an environment with a certain maximum allowable response delay, that is, the reward value obtained by the intelligent agent at this time is the maximum. Through continuous interaction with the environment during the training process, the intelligent agent eventually reaches an optimal strategy π that can be found in each environment. θ Maximize the current reward.

[0127] Virtual power plant assessment process Figure 7 As shown in the figure, the evaluation process is mainly divided into training process and testing process. The training process uses the terminal historical adjustable capacity and network status after edge cloud computing as the training set, and trains the neural network parameters so that each t allow The total amount of adjustable capacity of the flexible load is the largest, and the output neural network parameters are saved. This process usually takes a long time, and the historical data set is continuously updated by rolling training. The test process is a real-time evaluation process, and its input test set is the current terminal adjustable capacity and historical status, which can instantly evaluate the maximum adjustable capacity in the current environment.

[0128] Specifically, the adjustable potential evaluation optimization process can be regarded as a Markov decision process. In an interaction between the agent and the environment, at time t, the environment is in state s t , according to the current strategy π, the agent performs action a t And interact with the environment to get reward r t , which indicates the quality of the current action, and the environment moves to the next state s t+1 The above process is repeated. The key elements of the Markov decision process in the present invention are defined as follows.

[0129] 1) State space:

[0130]

[0131] The state space represents the environmental state of the system, which includes the available response capacity of the demand response terminals within the coverage area of ​​the regional distribution substation, the maximum allowed response delay, and the distribution network state uploaded by the collection module.

[0132] 2) Action Space:

[0133] A t ={x i (t),x ij (t)|i=1:I,j=1:J} (14)

[0134] The action space represents the variables in the optimization process, including the edge node selection x i and BS node selects x ij , is a discrete action space. This action space is the result of inputting the evaluation system state space into the policy function learned by the agent.

[0135] 3) Rewards:

[0136]

[0137] The total adjustable capacity F is used as the reward value obtained by the agent after one learning, and the average value of Q Monte Carlo simulations is sampled as the output to mitigate the fluctuation of the communication environment.

[0138] During the training process, the cloud node uses historical data within a certain sliding time window to train the model and obtain the cloud optimization decision model. During the real-time decision-making process, the cloud node inputs the real-time virtual power plant state space parameters into the model, uses the trained decision model to obtain the real-time optimization variables, and sends the results to the edge node. The edge node piecewise linearizes the objective function, converts the problem into a mixed integer linear programming problem, and then uses the solver to solve it.

[0139] In order to verify the effectiveness and rationality of the comprehensive evaluation method for the adjustable potential of virtual power plants based on the cloud-edge collaborative architecture taking into account both communication and action delays, the present invention adopts the following method: Figure 8 The IEEE 33-node distribution network system shown in the figure is used for demand response evaluation, with a voltage level of 12.66 kV. One cloud node, three edge nodes and 33 distribution substations including BS are set up, and the nodes in the power backbone network are connected by routers. Each BS covers 100 user demand response terminals. The BS is located in three different areas, where nodes 0 to 11 are residential areas, including interruptible loads, temperature control loads and electric vehicles; nodes 12 to 24 are commercial areas, mainly temperature control loads; nodes 25 to 32 are industrial areas, mainly industrial loads.

[0140] Typical daily adjustable potential curves of four loads at the demand response terminal are as follows: Fig. 9As shown. The theoretical maximum adjustable capacity and network status parameter matrix calculated from the response terminal loads in a certain place within ten days are selected as the training set of the algorithm. For the evaluation of the adjustable potential of the virtual power plant, we set five scenario modes. Mode 1 considers the communication system delay, equipment action time and communication reliability. Mode 2 only considers the equipment action time. Mode 3 is the case where the system is all residential areas. Mode 4 is the case where the system is all commercial areas. Mode 5 is the case where the system is all industrial areas. The relevant simulation parameter settings are shown in the following table.

[0141]

[0142] Optimization results:

[0143] This paper uses the PPO algorithm and the DDPG algorithm for example simulation. The reward curve comparison of the training process is as follows: Fig.10 As shown, the reward value of the PPO training process gradually increases and tends to be stable, and converges at about 11,000 steps, indicating that the agent has found a strategy to maximize the total amount of load adjustable capacity during the training process. As a traditional DRL algorithm, the DDPG algorithm tends to converge at around 12,000 steps, with large fluctuations in the convergence process, and the reward convergence performance of the training process is weaker than that of PPO, indicating that in the simulation environment set by the present invention, the stability and convergence of PPO are better than those of the traditional DRL algorithm. Using the adjustable capacity of the eleventh-day response terminal as the test set input, the RMSE value of the deviation between the PPO online evaluation result and the true value after denormalization is 0.129, and the evaluation result is relatively accurate.

[0144] The evaluation results of the daily adjustable potential of the central cloud virtual power plant considering time delay are as follows: Fig.11As shown in the figure, the total adjustable load in the area within the maximum allowable response delay range [0,500ms] with a step size of 5ms at any time of the day is shown. Mode 1 considers the communication system delay, equipment action time and communication reliability, while mode 2 only considers the equipment action time. The total adjustable capacity of mode 1 is 20.3% less than that of mode 2. This is because mode 2 lacks the communication delay part of the allowable response delay, and the allowable response delay is entirely provided by the equipment action delay, resulting in an error. When the allowable response delay is small, due to the existence of communication delay, mode 1 can only schedule terminals close to the server node, and there are fewer terminals that can respond, with small adjustable potential, and in a slow growth period. Mode 2 has the potential to schedule most terminals in advance, mainly including temperature control loads and interruptible loads with fast action speed, resulting in the largest error. As the allowable response delay increases, more demand response terminals can be added. After 100ms, due to the existence of electric vehicle loads, their high ramp rate causes the adjustable capacity to enter a period of substantial growth, and the error of mode 2 is also relatively large. The growth of adjustable capacity after 200ms is mainly provided by temperature-controlled industrial load and industrial load, which is determined by the nature of their long ramp-up time. It can be foreseen that a few seconds after the dispatching instruction is issued, the adjustable capacity can still increase with the growth of industrial load response capacity.

[0145] like Fig.12 The figure shows the proportion of each part of the communication delay of each BS node at 12:00, when the allowed response delay is 100ms. At this time, all responding terminals participate in the scheduling. The wired delay is related to the distance between the BS node and the edge node. The closer the node is, the smaller the proportion of the wired delay to the total communication delay, and the larger the schedulable action time. Since all terminals under each BS node are scheduled, the wireless delay and the computing delay are close to each other.

[0146] Modes 3, 4, and 5 are set to the situation where all the node distribution network systems are residential areas, commercial areas, and industrial areas, respectively. Fig.13 The figure shows the adjustable potential of the virtual power plant in each mode at 12:00. The shaded part indicates the range of adjustable potential under Monte Carlo sampling when the packet loss rate is 1% and the access network is lightly loaded. The volatility of the adjustable potential will be slightly higher when the delay is low. By comparing the modes, it can be found that the adjustable load potential in general within the 300ms allowable response delay is 23.4% and 12.7% higher than that of modes 3 and 5, respectively, and 22.3% lower than that of mode 4.

[0147] Sensitivity analysis:

[0148] To analyze the deviation of load scalability potential estimation due to packet loss during transmission, Fig.14 Shows the allowed response delay t allowThe maximum load adjustable potential changes with packet loss rate at 50ms, 100ms, 150ms, and 200ms. The shaded area is the distribution range of the adjustable potential after 100 Monte Carlo samplings, and the sample average is represented by a line. allow =50ms, the fluctuation range of the sampling results is large. This is because the adjustable capacity mainly depends on the communication delay. The packet loss rate increases the fluctuation of the communication delay. When there are fewer terminals that can be scheduled, even small fluctuations will increase the deviation of the evaluation. As the allowed response delay increases, the proportion of communication delay in the total allowed delay decreases, and the evaluation fluctuations generated by it also gradually decrease. Due to the existence of retransmission delay, when the packet loss rate increases, the adjustable capacity tends to decrease slowly.

[0149] like Fig.15 The figure shows the relationship between the wireless transmission delay and the load rate in the access network. The Monte Carlo sampling results are represented by scattered points. The change of the access network load rate has little effect on the overall delay. When the load rate increases, the delay may double in extreme scenarios, affecting the accuracy of the evaluation results.

[0150] To sum up, the comprehensive evaluation method for the adjustable potential of virtual power plants based on the cloud-edge collaborative architecture and taking into account both communication and action delays proposed in the present invention can accurately consider the impact of the communication system and resource action time to comprehensively evaluate the adjustable potential of virtual power plants, and can provide a decision-making basis for virtual power plants to participate in the primary, secondary frequency regulation and peak regulation of the power system, which is beneficial to supporting the frequency stability and power balance of the power system.

[0151] The computer device of the present invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0152] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program implements the steps of the above method when executed by a processor.

Claims

1. A comprehensive evaluation method for the adjustable potential of a virtual power plant taking into account communication and action delays, characterized in that: The following steps are involved: According to the virtual power plant cloud-edge collaborative network system framework, the communication system delay model, equipment action time model and communication reliability model are constructed respectively; According to the virtual power plant cloud-edge collaborative network system framework, communication system delay model, equipment action time model and communication reliability model, the KSP algorithm is used to obtain the optimal path and backup path between the source node m and the destination node n, and determine the total transmission delay t between the source node m and the destination node n during data transmission. d , build a maximum adjustable potential evaluation model for virtual power plants based on timed delay step search; The PPO algorithm is used to optimize the maximum adjustable potential evaluation model of the virtual power plant, and an online evaluation of the adjustable potential of the virtual power plant is obtained.

2. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 1 is characterized in that: The virtual power plant cloud-edge collaborative network system framework includes a cloud layer, an edge layer and a response terminal; the cloud layer is used to centrally integrate the computing resources of many servers and is responsible for processing global data services, including comprehensive evaluation, prediction of the aggregated adjustable capacity of virtual power plants, decomposition of scheduling instructions to each edge node, and overall communication optimization control of the cloud-edge-end system; the edge layer is used to calculate the aggregated load resource model at the edge nodes, and the edge nodes of the edge layer are set close to the user side.

3. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 1 is characterized in that: The communication system delay model is as follows: Wired transmission delay between the central cloud and edge nodes t t i ran for Among them, x i As a binary variable, x i =1 means that there is data packet transmission between cloud node and edge node i; otherwise x i =0;L i represents the size of the downlink data packet of the i-th node; r i is the wired transmission rate between the central cloud and the edge nodes; b i represents the channel bandwidth between cloud nodes, P i represents the data packet transmission power, G i represents the channel gain, and G i =d i -α , d i represents the distance between the ith node and the central cloud node, -α is the signal attenuation factor, and s represents the transmission noise; Wired propagation delay between cloud nodes and edge nodes for Where v is the propagation speed of the optical signal in the optical fiber; Downlink data delay between cloud edge and edge node t i for Among them, N h Indicates the number of routing hops, t f Indicates the processing delay of a single device; The jth edge node calculates the delay for Among them, f j is the computing power of the jth edge node, e j The number of CPU cycles required to calculate the task; Wireless download delay t p for in, represents user p in the jth region node j The queue position where it is located, T is the smallest indivisible period in wireless transmission; Wireless delay t p Using Pareto distribution to describe, the probability distribution function P is Among them, τ0 represents the minimum delay between nodes, γ is a parameter related to the load rate, and the cloud server calculates the delay is a constant, t p Wireless delay is generated in the process of transmitting from the base station to the demand response terminal.

4. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 3 is characterized in that: The communication reliability model is as follows: The reliability S between nodes m and n mn for Among them, N mn represents the number of links between m and n; S(e l ) represents link e l The reliability of the link. Information component failure, transmission packet loss, and bit error will affect the reliability of the link. mn The upper limit of the route hop count is set to 5. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 4 is characterized in that: The KSP algorithm is used to obtain the optimal path and backup path between the source node m and the destination node n. The deviation link weight of the transmission path between m and n is as follows: Among them, W mn is the link weight between nodes m and n, t mn is the delay between nodes m and n, c l represents the packet loss coefficient, Indicates rounding up, t s is the delay caused by packet loss and bit error, set to a constant, and d is a random number generated between 0 and 1.

6. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 5 is characterized in that: The total transmission delay t between the source node m and the destination node n during data transmission d for Among them, t ij The dispatching instruction sent to the edge node generates wired delay through the wireless base station, t a is the action time of the device, t s The delay caused by packet loss and bit errors.

7. The comprehensive evaluation method for adjustable potential of virtual power plants taking into account communication and action delays according to claim 6 is characterized in that: The maximum adjustable potential evaluation model of virtual power plant based on timed delay step search is: Among them, F is the total adjustable capacity of the central cloud virtual power plant, Indicates that the pth wireless base station is within the coverage area of ​​the jth wireless base station. j The maximum adjustable capacity of a demand response terminal, t allow is the maximum allowable response delay, x ij It is the path selection result between edge nodes and BS nodes in the load scheduling communication link.

8. A comprehensive evaluation system for the adjustable potential of a virtual power plant taking into account communication and action delays, characterized in that: include The model building module is used to build a communication system delay model, a device action time model and a communication reliability model according to the virtual power plant cloud-edge collaborative network system framework; it is used to use the KSP algorithm to obtain the optimal path and backup path between the source node m and the destination node n according to the virtual power plant cloud-edge collaborative network system framework, the communication system delay model, the device action time model and the communication reliability model, and determine the total transmission delay t between the source node m and the destination node n during the data transmission process. d , build a maximum adjustable potential evaluation model for virtual power plants based on timed delay step search; The strategy optimization evaluation module is used to determine the optimal strategy for the maximum adjustable potential evaluation model of the virtual power plant through the PPO algorithm, and obtain an online evaluation of the adjustable potential of the virtual power plant.

9. A computer 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 computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

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

Images