Whole-county photovoltaic data sharing value-added system and method based on 5G and federated learning

Abstract

The present disclosure provides a whole-county photovoltaic data sharing value-added system and method based on 5G and federal learning, which effectively balances the requirements and costs of power safety protection, uses a 5G power production control private network, collects photovoltaic data from photovoltaic terminals to a 5G power all-in-one fusion terminal, uploads the collected data to a power dispatch center by the all-in-one fusion terminal, and realizes the monitoring and regulation of the photovoltaic terminal by the power dispatch center; meanwhile, considering the communication requirements and data privacy of each demand side, the power grid company unifies the construction of a photovoltaic monitoring data sharing and value-added platform, establishes a data transmission channel between the data sharing and value-added platform and the management platform of each party, and realizes the multi-party sharing of photovoltaic data.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of photovoltaic data sharing, and particularly relates to a whole-county photovoltaic data sharing value-added system and method based on 5G and federated learning. BACKGROUND

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] In order to better realize the consumption of photovoltaic new energy, guarantee the power supply quality and operation safety of the power grid, and actively promote the 380V / 220V / 10kV distributed photovoltaic access transformation under the current situation, realize the access of all distributed photovoltaics to the municipal dispatching automation system, deploy the AGC (Automatic Generation Control) function to realize the smooth adjustment and control of power, and realize the observability, manageability, adjustability and controllability of distributed photovoltaic business.

[0004] The inventors found that in the whole-county (city, district) distributed photovoltaic scenario, there are currently three outstanding problems:

[0005] (1) The equipment of manufacturers is separate, the types are many, the concentration is low, and the operation and maintenance cooperation is difficult. The scene needs to complete the vertical encryption of photovoltaic regulation data, photovoltaic terminal state monitoring and remote control, AGC regulation, time synchronization and other functions, which causes the need to deploy multiple types of equipment (involving different professionals, even different manufacturers) on site, and often needs to coordinate multiple professionals on site during construction and testing, resulting in low operation and maintenance management efficiency.

[0006] (2) The traditional communication has high investment and cost, and must meet the safety protection requirements of regulation business. According to the requirements of power safety protection, production regulation business and management business need to be physically isolated. The communication problem of whole-county (city, district) photovoltaics can be divided into LAN side and WAN side. LAN side networking is the communication between the terminals of inverters and other terminals to the power local control terminal of the booster station. Traditionally, wired RS485 signal lines are connected in series and deployed in parallel with the high-voltage line, which is easy to be disturbed. According to the reliability specification, it should be spaced more than 10 cm. If it is not in the same slot according to the specification, it will increase the engineering cost. The HPLC (low-voltage broadband power line carrier) and micro-power wireless communication schemes used by traditional 380V photovoltaics respectively have the problems of being unable to cross inverters and unreliable communication. WAN side networking is the communication between the booster station as the convergence point and the dispatching automation master station. At present, a large number of newly-built demonstration points do not have optical fiber coverage. The cost of newly-built optical fiber communication network is high, and it needs to be safely isolated. The time delay and safety of traditional 4G wide area wireless network cannot meet the requirements of power dispatching. In addition, if each inverter is connected in a separate way, a security encryption device needs to be configured, and whether wired or wireless, the cost is high.

[0007] (3) Lack of mature and unified management mode or standard. The power dispatch center needs to monitor and control the inverter and the like, while the photovoltaic investment operator, inverter manufacturer, and telecom operator all have their own management platforms, and also need to manage and maintain their equipment assets. However, the current photovoltaic inverter does not have the ability to communicate with multiple parties at the same time, even if multiple communication ports are integrated, the manufacturing complexity and cost of the photovoltaic inverter will inevitably increase; all parties need more global data for power output prediction, equipment health assessment, and equipment improvement, but the conventional mode cannot obtain multi-dimensional data such as multi-provincial and multi-city, multi-investment, and multi-vendor, making it difficult to increase the value of data.

[0008] In summary, in the face of large-scale distributed photovoltaic grid connection, how to efficiently and safely operate the distributed photovoltaic energy system directly affects the economic benefit and safety and stability of the entire power system. The current industry lacks low-cost communication solutions and related business management modes, and it is inevitable to explore an effective management and operation mode that is safe, efficient, and low-cost and takes into account the needs of all parties in the context of distributed photovoltaic in the whole county (city, district). SUMMARY

[0009] To solve the above problems, the present disclosure provides a whole-county photovoltaic data sharing and value-added system and method based on 5G and federated learning, which effectively balances the requirements and costs of power safety protection. The photovoltaic data is collected from photovoltaic terminals to a 5G power all-in-one fusion terminal using a 5G power production control private network, and the all-in-one fusion terminal uploads the collected data to a power dispatch center to realize monitoring and control of the photovoltaic terminal by the power dispatch center. At the same time, considering the communication needs and data privacy of all parties, the power grid company constructs a photovoltaic monitoring data sharing and value-added platform, establishes a data transmission channel between the data sharing and value-added platform and the management platforms of all parties, and realizes multi-party sharing of photovoltaic data.

[0010] According to a first aspect of the embodiments of the present disclosure, a whole-county photovoltaic data sharing and value-added system based on 5G and federated learning is provided, comprising:

[0011] a photovoltaic grid-connected inverter for connecting the power generated by distributed photovoltaic to the power grid, wherein the photovoltaic grid-connected inverter has a 5G communication stick integrated inside;

[0012] a 5G computing power base station for realizing multi-point to single-point data aggregation of the photovoltaic grid-connected inverter to a 5G power all-in-one fusion terminal, wherein the 5G computing power base station is a 5G computing power single board deployed on a baseband processing unit of a 5G base station;

[0013] 5G power all-in-one fusion terminal, which is used for data collection, vertical encryption, remote control, AGC automatic power control, AVC automatic voltage control and time service of photovoltaic grid-connected inverter through integrated terminal; and the encrypted photovoltaic data is uploaded to the city power dispatching center;

[0014] The power dispatching center is used for receiving the photovoltaic data uploaded by the 5G power all-in-one fusion terminal, and transmitting the photovoltaic data to the provincial management information district server by using the power optical transmission network.

[0015] The data sharing and value-added platform includes an original data platform for data sharing and a federal learning platform for data value-added, the original data platform is used for obtaining the provincial photovoltaic data from the provincial management information district server, and transmitting the photovoltaic data to different demanders in a classified manner; and the federal learning platform is used for obtaining the federal learning results required by different demand methods by using the obtained photovoltaic data, and feeding back to each demander.

[0016] Further, after the distributed photovoltaic is accessed to the power grid, the photovoltaic data is collected by the photovoltaic grid-connected inverter to the 5G power all-in-one fusion terminal through the 5G computing base station, and the all-in-one fusion terminal accesses to the power dispatching center, so that the photovoltaic data is transmitted to the power dispatching center.

[0017] Further, the power dispatching center is also used for issuing power regulation instructions to the AGC automatic power control module of the 5G power all-in-one fusion terminal based on the obtained photovoltaic data, and after the AGC automatic power control module completes the calculation, issuing telemetry instructions to the controller of the corresponding photovoltaic grid-connected inverter, so as to realize the power control of the distributed photovoltaic.

[0018] Further, the system constructs a 5G power production control private network based on the 5G computing base station, which is used for data communication of the whole system, the 5G power production control private network is based on the 5G production control district hard slicing network, based on the 5G computing base station, and adopts the city specially built user plane function, flexible Ethernet transmission slicing and air interface resource block reservation / 5G service quality identifier priority scheduling technology, meets the physical isolation requirements of distributed photovoltaic regulation and control business, realizes the multi-point to single-point 5G local data aggregation function of the photovoltaic grid-connected inverter to the 5G all-in-one fusion terminal, and realizes the functions of collecting and monitoring the real-time data of the photovoltaic of the whole province in each city of the province.

[0019] Further, the 5G power production control private network has the end-to-end management and monitoring capability of slicing, the provincial power company separates the photovoltaic data according to different demanders, and transmits to each demander, so as to meet the conventional operation monitoring requirements of each demander.

[0020] Further, the 5G all-in-one fusion terminal is integrated with a longitudinal encryption chip, adopts an SM2 encryption algorithm, and realizes encrypted transmission of photovoltaic terminal data connected to 5G.

[0021] Further, the federal learning platform is composed of a local federal learning substation and a federal learning master station, photovoltaic data required by each demander is learned and trained on the corresponding local federal learning substation, and a parameter model obtained by local training is uploaded to the federal learning master station; the federal learning master station obtains a global parameter model by aggregating local parameter models of each demander for multiple times; the federal learning master station distributes the global parameter model to each local federal learning substation, the local federal learning substation trains the global parameter model, and finally obtains a federal learning result of the provincial photovoltaic data, and each demander obtains the federal learning result from the local federal learning substation.

[0022] Further, the federal learning result includes but is not limited to photovoltaic output prediction and equipment health index.

[0023] Further, the demander includes an investment operator, an equipment supplier, and a telecom operator.

[0024] According to a second aspect of the embodiments of the present disclosure, a whole-county photovoltaic data sharing and value-added method based on 5G and federal learning is provided, which is based on the above-mentioned whole-county photovoltaic data sharing and value-added system based on 5G and federal learning, and the method comprises the following steps:

[0025] Based on the 5G power all-in-one fusion terminal, distributed photovoltaic data is collected and transmitted to a power dispatching center;

[0026] The power dispatching center monitors and regulates the distributed photovoltaic based on the obtained photovoltaic data; meanwhile, based on the data sharing and value-added platform, data transmission between the provincial distributed photovoltaic data and management platforms of each party is realized, and multi-party sharing and data value-added of the photovoltaic data are realized.

[0027] Compared with the prior art, the present disclosure has the following beneficial effects:

[0028] (1) The present disclosure provides a 5G and federated learning based whole county photovoltaic data sharing and value added system and method. Compared with the existing photovoltaic grid connected communication scheme, the scheme described in the present disclosure is based on the characteristics of the whole county (city, district) distributed photovoltaic scene, and a set of end-to-end innovative management mode suitable for the distributed photovoltaic business scene is designed from the terminal, network, platform and application. Considering the requirements and costs of power safety protection, the 5G power production control private network is used to collect photovoltaic data from photovoltaic terminals to 5G power all-in-one fusion terminals, and the all-in-one fusion terminals upload the collected data to the power dispatching center to realize the monitoring and regulation of the photovoltaic terminal by the power dispatching center. At the same time, considering the communication needs and data privacy of each demand side, the scheme described in the present disclosure realizes the multi-party sharing of photovoltaic data by constructing a photovoltaic monitoring data sharing and value added platform and establishing a data transmission channel between the data sharing and value added platform and the management platform of each party.

[0029] (2) The scheme described in the present disclosure fuses the functions of data acquisition, vertical encryption, remote control, AGC automatic power control, AVC, time service, etc. of photovoltaic inverter and other terminals through 5G power all-in-one fusion terminal equipment, so that the originally scattered device functions are integrated into one, and one device can realize all functions, reducing the difficulty of device deployment and operation and maintenance.

[0030] The advantages of the additional aspects of the present disclosure will be partially given in the following description, partially will become obvious from the following description, or will be known by the practice of the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0031] The drawings accompanying the specification of the present disclosure serve to provide further understanding of the present disclosure, and the illustrative embodiments of the present disclosure and their descriptions serve to explain the present disclosure, and do not constitute an improper limitation on the present disclosure.

[0032] Figure 1 The basic structure schematic diagram of the 5G and federated learning based whole county photovoltaic data sharing and value added system described in the embodiments of the present disclosure is shown in the figure;

[0033] Figure 2 The whole framework structure schematic diagram of the 5G and federated learning based whole county photovoltaic data sharing and value added system described in the embodiments of the present disclosure is shown in the figure;

[0034] Figure 3 The power 5G production control slice architecture schematic diagram described in the embodiments of the present disclosure is shown in the figure;

[0035] Figure 4 The power 5G distributed business communication architecture schematic diagram described in the embodiments of the present disclosure is shown in the figure;

[0036] Figure 5 The traditional base station mode schematic diagram described in the embodiments of the present disclosure is shown in the figure;

[0037] Figure 6 A 5G computing power base station mode schematic diagram described in the embodiments of the present disclosure;

[0038] Figure 7 A traditional whole-county photovoltaic communication network connection schematic diagram described in the embodiments of the present disclosure;

[0039] Figure 8 A traditional whole-county photovoltaic commercial management mode schematic diagram described in the embodiments of the present disclosure;

[0040] Figure 9 A whole-county photovoltaic commercial management mode schematic diagram of a whole-county photovoltaic data sharing value-added system based on 5G and federated learning described in the embodiments of the present disclosure. DETAILED DESCRIPTION

[0041] The present disclosure will be further described below in conjunction with the accompanying drawings and embodiments.

[0042] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure belongs.

[0043] It should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit exemplary embodiments according to the present disclosure. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should also be understood that when the terms "comprise" and / or "include" are used in the specification, there is a presence of a feature, step, operation, device, component, and / or combinations thereof.

[0044] The embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0045] Embodiment one:

[0046] The purpose of this embodiment is to provide a whole-county photovoltaic data sharing value-added system based on 5G and federated learning.

[0047] As shown in Figure 1 and Figure 2 A whole-county photovoltaic data sharing value-added system based on 5G and federated learning, comprising:

[0048] A photovoltaic grid-connected inverter for connecting distributed photovoltaic generated power to the grid, wherein the photovoltaic grid-connected inverter has a 5G communication stick integrated inside;

[0049] 5G computing power base station for implementing the multi-point to single-point data aggregation of the photovoltaic grid-connected inverter to the 5G power all-in-one fusion terminal, wherein the 5G computing power base station is a 5G computing power single board deployed on a baseband processing unit of a 5G base station;

[0050] 5G power all-in-one fusion terminal for data acquisition, vertical encryption, remote control, AGC automatic power control, AVC automatic voltage control, and time service of the photovoltaic grid-connected inverter through an integrated terminal; and uploading the encrypted photovoltaic data to a prefecture-level power dispatching center;

[0051] Power dispatching center for receiving the photovoltaic data uploaded by the 5G power all-in-one fusion terminal, and sending the photovoltaic data to a provincial management information district server through a power optical transmission network;

[0052] Data sharing and value-added platform including an original data platform for data sharing and a federal learning platform for data value-added, the original data platform is used to obtain the provincial photovoltaic data from the provincial management information district server, and transmit the photovoltaic data to different demand parties in a classified manner; the federal learning platform is used to obtain the federal learning results required by different demand methods using the obtained photovoltaic data, and feedback to each demand party.

[0053] Further, after the distributed photovoltaic is accessed to the power grid, the photovoltaic data is collected by the photovoltaic grid-connected inverter to the 5G power all-in-one fusion terminal through the 5G computing power base station, and the all-in-one fusion terminal accesses to the power dispatching center, so that the photovoltaic data is transmitted to the power dispatching center.

[0054] Further, the power dispatching center is also used to issue a power regulation instruction to an AGC automatic power control module of the 5G power all-in-one fusion terminal based on the obtained photovoltaic data, and after the AGC automatic power control module completes the calculation, issues a telemetry instruction to a controller of the corresponding photovoltaic grid-connected inverter, so as to realize the power control of the distributed photovoltaic.

[0055] Further, the system constructs a 5G power production control private network based on the 5G computing power base station, which is used for data communication of the entire system, the 5G power production control private network is based on a 5G production control district hard slicing network, based on the 5G computing power base station, and adopts a city specially built user plane function, a flexible Ethernet transmission slice, and an air interface resource block reservation / 5G service quality identifier priority scheduling technology, meets the physical isolation requirements of the distributed photovoltaic regulation and control business, realizes the multi-point to single-point 5G local data aggregation function of the photovoltaic grid-connected inverter to the 5G all-in-one fusion terminal, and realizes the functions of collecting and monitoring the real-time data of the photovoltaic in the whole province of each city in the province in units of cities.

[0056] Further, the 5G power production control private network has end-to-end management and monitoring capabilities of slices. The provincial power company separates the photovoltaic data according to different demanders and transmits the data to the demanders, thereby meeting the conventional operation monitoring requirements of the demanders.

[0057] Further, the 5G multi-in-one fusion terminal integrates a longitudinal encryption chip and adopts an SM2 encryption algorithm to realize encrypted transmission of photovoltaic terminal data under 5G.

[0058] Further, the federal learning platform is composed of a local federal learning substation and a federal learning master station. The photovoltaic data required by the demanders are trained on the corresponding local federal learning substation, and the parameter model obtained by the local training is uploaded to the federal learning master station. The federal learning master station aggregates the local parameter models of the demanders for multiple times to obtain a global parameter model. The federal learning master station distributes the global parameter model to each local federal learning substation, and the local federal learning substation trains the global parameter model to finally obtain the federal learning result of the provincial photovoltaic data. The demanders obtain the federal learning result from the local federal learning substation.

[0059] Further, the federal learning result includes but is not limited to photovoltaic output prediction and equipment health index.

[0060] Further, the demanders include investment operators, equipment suppliers and telecom operators.

[0061] Specifically, in order to facilitate understanding, the scheme described in the embodiment is described in detail below in combination with the drawings:

[0062] In order to solve the problems in the prior art, the embodiment provides an entire county photovoltaic data sharing and value-added system based on 5G and federal learning, and the main technical concept is as follows:

[0063] Based on the characteristics of distributed photovoltaic scene in the whole county (city, district), a set of end-to-end innovation management mode suitable for distributed photovoltaic business scene is designed from terminal, network, platform and application. Considering the requirements and cost of power safety protection, the 5G power production control special network is used to collect photovoltaic data from photovoltaic terminal to 5G power all-in-one fusion terminal, and the all-in-one fusion terminal uploads the collected data to the power dispatching center to realize the monitoring and regulation of photovoltaic terminal by the power dispatching center. Considering the communication needs and data privacy of each demand side, the power grid company unifies to build photovoltaic monitoring data sharing and value-added platform, establishes data transmission channel between data sharing and value-added platform and management platform of each party, and realizes multi-party sharing of photovoltaic data. The mode uses 5G power all-in-one fusion terminal, which realizes all functions of photovoltaic inverter and other terminals with one device, solves the problems of difficult equipment deployment and low operation efficiency; 5G power production control special network is used for data communication, based on 5G computing power base station local networking, eliminating transmission detours, improving data processing efficiency, at the same time, 5G end-to-end hard slicing provides safe and reliable communication guarantee, meets the requirements of power safety protection, solves the problems of large investment and high cost of traditional communication; photovoltaic monitoring data sharing and value-added platform is used to realize data sharing and value-added, and a win-win business management mode is proposed to ensure data privacy and safety, and solve the problems of difficult multi-party management and data value-added. Specifically, the main technologies used in the scheme of the embodiment are described in detail as follows:

[0064] (1) 5G power all-in-one fusion terminal / 5G communication stick scheme

[0065] This scheme follows the latest version of national standard and power industry standard. The standard specifies the technical requirements of 5G wireless communication accessories, remote control, longitudinal encryption, etc., including structure, working environment, basic transmission characteristics, electrical safety and electromagnetic compatibility requirements. 5G power all-in-one fusion terminal device integrates data acquisition, longitudinal encryption, remote control, AGC automatic power control, AVC, time service and other functions of photovoltaic inverter and other terminals, which makes the originally scattered device functions zero into a whole, and one device can realize all functions, reducing the difficulty of equipment deployment and operation and maintenance. 5G communication stick supports 485 interface and is integrated into photovoltaic terminal inverter to form 5G transmission link from each photovoltaic inverter to 5G power all-in-one fusion terminal.

[0066] After the distributed photovoltaic access to the power grid, photovoltaic data information is collected by photovoltaic terminal inverters to the 5G power all-in-one fusion terminal in a "multi-point to one-point" manner through 5G, and the all-in-one fusion terminal accesses the power dispatch center, so that the data is landed on the power dispatch center. The dispatch master station can issue power control instructions to AGC according to photovoltaic data, and the AGC component calculates and issues telemetry instructions to each set of inverter controller. At the same time, the 5G power all-in-one fusion terminal has containerized app management function, and the later product also plans to add AVC function and edge computing capability, and through 5G precise timing function, another clock source guarantee is provided for local booster station, realizing the earth-sky backup of power clock synchronization network.

[0067] (II) 5G power production control private network solution based on 5G computing power base station

[0068] Based on the existing 5G production control district hard slicing network, the 5G computing power single board is deployed on the base station BBU (Building Base band Unit, baseband processing unit) to upgrade to a 5G computing power base station. The "city special UPF (User Plane Function, user plane function) + FlexE (Flexible Ethernet, flexible Ethernet) transmission slice + air interface RB (Resource Block, resource block) reservation / 5QI (5G QoS Identifier, 5G service quality identifier) priority scheduling" technical means are adopted to meet the physical isolation requirements of distributed photovoltaic regulation and control business, realize the multi-point to one-point 5G local data aggregation function from inverter to all-in-one fusion terminal, and realize the function of collecting and monitoring the real-time data of whole-provincial photovoltaic in each city and county of the province. Including the following characteristics:

[0069] Computing power single board dedicated to power: The southward (5G all-in-one fusion terminal-inverter) data of whole-county photovoltaic does not go out of the county, and the computing power resources of the base station are exclusively enjoyed by the county, which increases the isolation and security of data.

[0070] Utilization and performance improvement: cross-base-station terminal interconnection can be realized, that is, photovoltaic inverters can directly communicate with 5G power all-in-one fusion terminals through 5G base stations, eliminating the traffic detours of aggregation ring and UPF forwarding, and reducing local latency.

[0071] Reliability improvement: local shunting reduces the number of links through which local business flows, and the interworking configuration between UPF and terminal is reserved. When the local shunting of the computing power single board fails, communication can be switched to the city UPF in time.

[0072] Management openness: it can be managed through the operator base station network management, and it supports northbound API interface, so that the power can arrange business links through operator capability opening.

[0073] Further, as shown in Figure 5 and Figure 6 The 5G computing power base station has the following advantages compared with the 5G traditional base station:

[0074] (1) It has greater throughput processing capacity and can meet the aggregation and uploading of all photovoltaic data in the county.

[0075] (2) In the 5G traditional base station mode, the data transmission path between terminals is: inverter terminal-5G traditional base station-city UPF-5G traditional base station-5G multi-in-one fusion terminal. In the 5G computing power base station mode, the data transmission path between terminals is: inverter terminal-5G traditional base station-5G computing power base station-5G traditional base station-5G multi-in-one fusion terminal. Generally, one city UPF is deployed in one city, and one 5G computing power base station is deployed in one county. In the 5G computing power base station mode, data passes through the nearest county-level computing power base station, rather than passing through the city-level UPF, thereby reducing the detour of traffic and reducing the transmission delay. The process is shown in the following diagram.

[0076] (3) The 5G computing power base station supports northbound API interface and can be integrated into the operator base station network management. Through the operator capability opening, the power company can arrange business links by itself.

[0077] (4) The 5G computing power single board is dedicated to power companies, which can improve the information security isolation capability of 5G photovoltaic monitoring. In the 5G computing power base station mode, the southbound (5G multi-in-one fusion terminal-inverter) data of the whole county photovoltaic does not go out of the county, and the computing power resources of the base station are exclusively enjoyed by the county. At the same time of avoiding information flow detour, the data isolation is increased, and the security is higher.

[0078] (Three) Whole county (city, district) photovoltaic monitoring data sharing and value-added scheme based on federated learning

[0079] The photovoltaic data is uploaded to the dispatching automation system (power dispatching center) of each city power supply company after being encrypted by the 5G multi-in-one fusion terminal. The power dispatching automation system belongs to the production control area (area one and two), and the data collected by it cannot be directly transmitted to the outside. This scheme configures an isolation device between the dispatching automation system of each city and the server of the provincial company management information area (area three and four), uses the power optical transmission network to establish a data transmission channel from the city dispatching to the provincial company management information area, and transmits the photovoltaic data from the dispatching automation system D5000 server to the provincial power company area three and four server.

[0080] Among them, the production control area is divided into control area (safety area I, area I) and non-control area (safety area II, area II), the management information area is divided into production management area (safety area III, area III) and management information area (safety area IV, area IV), different safety areas have different safety protection requirements, among which the safety area I (area I) has the highest safety level, followed by the safety area II (area II), and the rest in turn. The terminals in different areas cannot access directly.

[0081] A data sharing and value-added platform (composed of an original data platform and a federated learning platform) is established in the provincial power company. Demanders such as investors, equipment manufacturers, and telecom operators obtain real-time photovoltaic monitoring data (photovoltaic operating status, fault information, etc.) from the original data platform in a "back-to-back" form, i.e., the demanders do not interact with each other and can only obtain data information belonging to themselves, and cannot obtain data information of other demanders, ensuring the safety and privacy of data transmission.

[0082] Due to the limited amount of original data obtained by each demander, the guidance precision of photovoltaic output prediction and equipment health indicators, which are highly concerned by each party, is insufficient. To improve the precision of data prediction and diagnosis, the provincial photovoltaic data needs to be integrated and trained, considering the need to ensure data privacy and not to disclose between demanders. This scheme combines AI artificial intelligence, big data, and other technologies, and uses the technical architecture of federated learning.

[0083] The federated learning platform is composed of local federated learning sub-stations and a federated learning master station. Taking investors and photovoltaic equipment manufacturers as demanders as an example, the data of each investor / photovoltaic equipment manufacturer is learned and trained on their respective local federated learning sub-stations (AI training sub-platform), and the parameter model obtained by local training is uploaded to the federated learning master station (AI training master platform). The federated learning master station aggregates the local parameter models of each party multiple times to obtain a global parameter model. The master station distributes the global parameter model to each local federated learning sub-station, which trains the global parameter model to finally obtain the federated learning results of the provincial photovoltaic data (photovoltaic output prediction, equipment health indicators, etc.). Each investor / photovoltaic equipment manufacturer can obtain the federated learning results from their respective local federated learning sub-stations.

[0084] (Four) Commercial management mode

[0085] The scheme based on the above data sharing and value-added platform proposes a business management mode. After the photovoltaic data is landed in the provincial power company, due to the end-to-end management and monitoring capability of the power 5G private network, the provincial power company can separate the data according to the needs of each investor operator, equipment manufacturer and other demand parties, and transmit it to each party to meet the regular operation monitoring needs of each party. In this process, each demand party can pay a certain data fee to the data platform operator. Since the power grid faces multiple demand parties, it is completely possible for a single demand party to achieve information cost lower than the wireless traffic cost of the operator. In addition, each demand party obtains the federated learning result from the above-mentioned respective local federated learning sub-station. In this process, each demand party can pay a certain data value-added fee to the data platform operator. The value-added data can be charged separately according to market demand, and on the premise of ensuring the data back-to-back security and privacy of each party, the data value-added that the traditional scheme cannot achieve is realized.

[0086] (Five) Distributed photovoltaic security scheme

[0087] The scheme centrally collects data of all access devices into a 5G all-in-one terminal. The 5G all-in-one fusion terminal provides unified access control for all access devices, physically reduces data dispersion and multiple sources, and configures identity authentication, data encryption, forward and reverse isolation and other functions on the business layer to meet the power safety protection requirements of all access devices. The 5G all-in-one terminal also integrates a vertical encryption chip, uses an SM2 encryption algorithm, and realizes encrypted transmission of photovoltaic terminal data connected under 5G. In the 5G private network transmission process, authentication between the terminal and the power dispatching center can also use national secret asymmetric cryptography algorithms and access control functions.

[0088] In addition, the 5G computing power base station introduced by the scheme also has a comprehensive security mechanism. Corresponding to the 10KV photovoltaic UE (User Equipment) to UE interworking scenario, it mainly applies to NE (Network Element) power exclusive, internal network isolation, UPF as NE link disaster recovery backup and other mechanisms.

[0089] Further, the following will explain the scheme described in the embodiment in detail through specific examples from the perspective of specific implementation:

[0090] The scheme described in the embodiment specifically includes the following steps:

[0091] Step one, deploy and configure photovoltaic inverter terminal and 5G all-in-one fusion terminal. In this embodiment, each household in a town in the county is equipped with a solar photovoltaic panel, and an average of 5 households of solar photovoltaic panels are connected to one photovoltaic grid-connected inverter. A 5G communication stick is integrated with the photovoltaic grid-connected inverter to enable the photovoltaic grid-connected inverter to have 5G communication function. One set of 5G all-in-one fusion terminal is configured at each corresponding photovoltaic booster station in each town.

[0092] Step two, establish a 5G power production control private network based on 5G computing power base station. In this embodiment, a 5G computing power single board is deployed on the BBU of a 5G base station in the county, and the base station is upgraded to a 5G computing power base station to realize the multi-point to single-point data aggregation function of the distributed photovoltaic all-in-one fusion terminal. One computing power single board can support 10G throughput processing capability, and other base stations on the backhaul network can use it for terminal interconnection. Other distributed power services (such as distribution network area protection and feeder automation) can also use the same computing power single board. The single board can also provide local photovoltaic data traffic offloading, and sink computing power to the county or even village level. A set of power dedicated UPF is deployed in the city power supply company, and a city dedicated UPF is formed to improve the reliability of photovoltaic regulation.

[0093] Step three, establish a photovoltaic data transmission channel between the city power supply company and the provincial power company. In this embodiment, an isolation device is configured on the D5000 server of the city dispatching automation system, and the photovoltaic data is transmitted from the D5000 server of the city dispatching automation system to the provincial power company three-fourth zone cloud platform through the power dedicated optical fiber transmission network.

[0094] Step four, establish a whole county (city, district) photovoltaic monitoring data sharing platform. In this embodiment, a raw data platform is established in the provincial power company, which contains photovoltaic data of all power generation users in the province. The platform can separate the photovoltaic data according to the needs of each demander such as investment operator and equipment supplier, and transmit it to the monitoring center of the corresponding demander.

[0095] Step five, establish a whole county (city, district) photovoltaic monitoring data value-added platform. In this embodiment, a federal learning platform is established in the provincial power company, which is composed of a local federal learning substation and a federal learning master station. The substation is responsible for local training and learning of photovoltaic data to obtain local parameter model, and the master station is responsible for aggregating and training each local parameter model to obtain global parameter model. The master station sends the global parameter model to each substation, and the substation trains and learns the global parameter model to finally obtain the federal learning result of the photovoltaic data, and sends the result to the monitoring center of the corresponding demander.

[0096] Figure 1It is the overall architecture schematic diagram of the scheme. The photovoltaic inverter and the 5G all-in-one fusion terminal communicate through the 5G base station. When the 5G computing power single board fails, the communication link can be switched to communicate through the city power special UPF. The 5G computing power base station transmits the photovoltaic data collected by the all-in-one fusion terminal to the dispatching automation system D5000 server through the power optical transmission network.

[0097] Figure 2 It is the overall architecture schematic diagram of the whole county photovoltaic data sharing system based on 5G+ federal learning. Through the power special optical transmission network, the photovoltaic data is transmitted from the dispatching automation system D5000 server to the provincial power company three-fourth cloud platform. The provincial power company three-fourth cloud platform is directly connected with the original data platform and sends the photovoltaic data (real-time running state, fault information) of all power generation users in the province to the original data platform. The process of data sharing and value-added platform is as follows: the original data platform separates the photovoltaic data of investment operator 1 and transmits it to the monitoring center of investment operator 1. The original data platform sends the photovoltaic data of investment operator 1 to federal learning substation 1 (each substation corresponds to each demand side, i.e. substation 1 corresponds to investment operator 1 and substation 2 corresponds to investment operator 2). Substation 1 obtains local parameter model 1 through data training and uploads it to the master station. The master station aggregates local model parameters 1, 2... N to obtain global parameter model. The master station sends the global model parameters to substation 1. Substation 1 obtains federal learning results through training and sends them to the monitoring center of investment operator 1.

[0098] Figure 3 It is the schematic diagram of the power 5G production control slice. The power 5G special network is in the mode of "public network special use". The operator divides the shared resources for the power grid company. The power special control plane network element is deployed in the operator provincial center machine room. The power special UPF is deployed in the city power supply company machine room. The power data stream is sent by the terminal, directly landed in the city power supply company through the power special UPF, and does not need to pass through the operator provincial center machine room again.

[0099] Figure 4 It is the schematic diagram of the power 5G distributed business communication architecture. The communication process between the photovoltaic inverter terminal and the 5G all-in-one fusion terminal can be completed through the 5G computing power base station and the 5G base station directly, without the need to forward through the power special UPF again. The communication process between the terminal and the power grid server needs to be completed through the 5G computing power base station and the power special UPF.

[0100] Figure 5 It is the schematic diagram of the traditional base station mode;

[0101] Figure 6 It is the schematic diagram of the 5G computing power base station mode;

[0102] Figure 7The traditional whole county photovoltaic communication network connection schematic diagram is shown. The LAN side networking is the communication between the terminal such as the inverter and the power local control terminal of the booster station, and the traditional one is usually wired RS485 signal line in series. The WAN side networking is the communication between the photovoltaic booster station as the convergence point and the dispatching automation master station.

[0103] Figure 8 The traditional whole county photovoltaic commercial management mode schematic diagram is shown. Each demand side communicates with the terminal separately, and each photovoltaic inverter needs three wireless communication cards and 4G / 5G communication sticks in the figure.

[0104] Figure 9 The whole county photovoltaic commercial management mode schematic diagram of the present scheme is shown. Each photovoltaic inverter only needs one wireless communication card and a 5G communication stick, and the data of all inverters is collected to the power grid dispatching center, and the data platform operator separates the data according to each demand side and transmits it to each demand side.

[0105] Embodiment two:

[0106] The purpose of the present embodiment is to provide a whole county photovoltaic data sharing and value-added method based on 5G and federated learning.

[0107] A whole county photovoltaic data sharing and value-added method based on 5G and federated learning, which is based on the above-mentioned whole county photovoltaic data sharing and value-added system based on 5G and federated learning, the method comprises:

[0108] Based on the 5G power all-in-one fusion terminal, the distributed photovoltaic data is collected and transmitted to the power dispatching center;

[0109] The power dispatching center monitors and regulates the distributed photovoltaic based on the obtained photovoltaic data; at the same time, based on the data sharing and value-added platform, the data transmission between the whole province distributed photovoltaic data and the management platform of each party is realized, and the multi-party sharing and data value-added of photovoltaic data is realized.

[0110] In more embodiments, the following are also provided:

[0111] An electronic device comprising a memory and a processor, and computer instructions stored on the memory and running on the processor, when the computer instructions are run by the processor, the method described in embodiment one is completed. For brevity, it will not be repeated here.

[0112] It should be understood that in the present embodiment, the processor can be a central processing unit CPU, and the processor can also be other general-purpose processors, digital signal processors DSPs, application-specific integrated circuits ASICs, ready-to-program gate arrays FPGAs or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0113] The memory can include read-only memory and random access memory, and provide the processor with instructions and data. A portion of the memory can also include non-volatile random access memory. For example, the memory can also store device type information.

[0114] A computer readable storage medium for storing computer instructions, when executed by a processor, completes the method described in embodiment one.

[0115] The method in embodiment one can be directly embodied as a hardware processor to complete, or be completed by a combination of hardware and software modules in the processor. The software modules can be located in the storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, register, etc. The storage medium is located in the memory, and the processor reads the information in the memory to complete the steps of the above method in combination with the hardware. To avoid repetition, it will not be described in detail here.

[0116] Those skilled in the art can realize that the units of the examples described in combination with the embodiments, i.e. the algorithm steps, can be realized in electronic hardware or in combination of computer software and electronic hardware. Whether the functions are executed in hardware or software mode depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the disclosure.

[0117] The 5G and federated learning-based whole-county photovoltaic data sharing and value-added system and method provided by the above embodiments can be realized, and has a broad application prospect.

[0118] The above only describes the preferred embodiments of the disclosure and is not intended to limit the disclosure. Those skilled in the art can make various modifications and changes to the disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the disclosure shall be included in the protection scope of the disclosure.

Claims

1. A 5G and federated learning based whole county photovoltaic data sharing value-added system, characterized in that, The application relates to a photovoltaic grid-connected inverter for connecting distributed photovoltaic generated power to a power grid, wherein a 5G communication rod is integrated in the photovoltaic grid-connected inverter. A 5G computing power base station is used to realize multi-point to single-point data convergence of the photovoltaic grid-connected inverter to a 5G power multi-in-one fusion terminal, wherein the 5G computing power base station is used to deploy a 5G computing power single board on a baseband processing unit of a 5G base station. A 5G power multi-in-one fusion terminal is used to collect data of the photovoltaic grid-connected inverter, vertically encrypt the data, remotely control the data, realize AGC automatic power control, AVC automatic voltage control and time service through an integrated terminal; and upload the encrypted photovoltaic data to a city power dispatching center. A power dispatching center is used to receive the photovoltaic data uploaded by the 5G power multi-in-one fusion terminal, and transmit the photovoltaic data to a provincial management information district server through a power optical transmission network; the power dispatching center is also used to issue a power control instruction to an AGC automatic power control module of the 5G power multi-in-one fusion terminal based on the obtained photovoltaic data, and the AGC automatic power control module issues a telemetry instruction to a controller of a corresponding photovoltaic grid-connected inverter after calculation, so that power control of distributed photovoltaic is realized. A data sharing and value-added platform comprises an original data platform for data sharing and a federal learning platform for data value-added, the original data platform is used to obtain provincial photovoltaic data from a provincial management information district server, and the photovoltaic data is classified and transmitted to different demanders; the federal learning platform is used to obtain a federal learning result required by different demanders by using the obtained photovoltaic data, and the federal learning result is fed back to each demander. The federal learning platform is composed of a local federal learning substation and a federal learning master station, photovoltaic data required by each demander is learned and trained on the corresponding local federal learning substation, and a parameter model obtained by local training is uploaded to the federal learning master station; the federal learning master station obtains a global parameter model by aggregating local parameter models of each demander for multiple times; the federal learning master station issues the global parameter model to each local federal learning substation, the local federal learning substation trains the global parameter model, and finally obtains a federal learning result of provincial photovoltaic data, and each demander obtains the federal learning result from the corresponding local federal learning substation. After the distributed photovoltaic is connected to the power grid, photovoltaic data is collected from the photovoltaic grid-connected inverter to the 5G power multi-in-one fusion terminal through the 5G computing power base station, the multi-in-one fusion terminal is connected to the power dispatching center, and the photovoltaic data is transmitted to the power dispatching center.

2. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 1, wherein, ​ 3. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 1, wherein, The system constructs a 5G power production control private network based on the 5G computing base station, which is used for data communication of the entire system, and the 5G power production control private network is based on a 5G production control large area hard slicing network, based on the 5G computing base station, and adopts a city-built user plane function, a flexible Ethernet transmission slice, and an air interface resource block reservation / 5G quality of service identifier priority scheduling technology to meet the physical isolation requirements of distributed photovoltaic regulation and control business, realize the multi-point-to-point 5G local data aggregation function of the photovoltaic grid-connected inverter to the 5G multi-in-one fusion terminal, and realize the function of collecting and monitoring the real-time data of the photovoltaic in the whole province in units of cities.

4. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 3, wherein, The 5G power production control private network has end-to-end management and monitoring capabilities of slicing, and the provincial power company separates the photovoltaic data according to different demand parties and transmits them to each demand party, thereby meeting the conventional operation monitoring requirements of each demand party.

5. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 1, wherein, The 5G multi-in-one fusion terminal integrates a longitudinal encryption chip and adopts an SM2 encryption algorithm to realize encrypted transmission of photovoltaic terminal data under 5G.

6. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 1, wherein, The federal learning result includes but is not limited to photovoltaic output prediction and equipment health index.

7. The 5G and federated learning based whole-county photovoltaic data sharing value-added system of claim 1, wherein, The demand parties include investment operators, equipment suppliers and telecom operators.

8. A whole-county photovoltaic data sharing value-added method based on 5G and federated learning, characterized in that, The method based on the 5G and federal learning-based whole-county photovoltaic data sharing and value-added system according to any one of claims 1-7, the method comprising: Based on the 5G power multi-in-one fusion terminal, the distributed photovoltaic data is collected and transmitted to the power dispatching center; The power dispatching center monitors and regulates the distributed photovoltaic based on the obtained photovoltaic data; at the same time, based on the data sharing and value-added platform, the data transmission of the whole-province distributed photovoltaic data and the management platforms of various parties is realized, and the multi-party sharing and data value-added of the photovoltaic data are realized.

Images