Intelligent collaborative power demand network method

Abstract

The invention relates to an intelligent collaborative power demand network method, and provides an intelligent collaborative theory-based power demand processing mechanism solution, which comprises the following steps of: firstly, designing a new data packet structure, adding demand identifiers and demand attribute information, and further clearly identifying and describing characteristics and requirements of different power grid business demands; more bases are provided for subsequent processing and forwarding; secondly, based on the characteristics of different service requirements in the scenes of transmission, transmission, transformation, distribution and utilization, decision-making numbering is carried out according to the communication requirements of the service requirements, and binding between the two is realized; and finally, setting a demand preference matching table in the intermediate equipment, obtaining a matching relationship between the business demand and the preference according to the matching model, and determining the preference of the demand by looking up the table. The optimal forwarding path is selected in combination with demand preference, sustainable control and reasonable resource allocation are realized, the efficiency of service transmission of the power Internet of Things can be effectively improved, and diversified demand perception in the power Internet of Things is supported.

Description

Technical Field

[0001] The present invention relates to the field of electric power, and more specifically to a smart collaborative electric power demand network method. Background Art

[0002] With the implementation of 5G, the era of the Internet of Things has officially begun. The continuous accumulation and upgrading of Internet of Things technologies have gradually improved and matured the industrial chain. In addition, driven by cyclical factors such as infrastructure construction, transformation of basic industries and consumption upgrades, fields and industries at different levels of development have alternately promoted the development of the Internet of Things, especially the power Internet of Things industry. The power Internet of Things will continuously integrate various acquisition and control sensors or controllers with perception and monitoring capabilities, as well as mobile communications, intelligent analysis and other technologies into all aspects of the power production process, and ultimately realize the upgrading of the traditional power industry to a new stage of intelligence. In terms of application form, the power Internet of Things application has the characteristics of real-time, automation, embedded (software), security, and information interoperability. The power Internet of Things market is developing rapidly, and it pays more attention to the needs of high reliability, low latency, and massive connections, which puts forward higher requirements for the realization of power automation.

[0003] The power Internet of Things is composed of the power backbone communication network, the real-time monitoring network of power plants and stations, and the power communication access network. Its needs can be divided into five aspects: power generation, transmission, transformation, distribution, and consumption, and each has its own characteristics in terms of perception of demand. The power Internet of Things needs to realize the real-time interconnection and dynamic interaction between the equipment involved in each link of the power system generation, transmission, transformation, distribution, and consumption. It has the dual attributes of network and control, and has the advantages of real-time, reliability, and inclusiveness. It is essentially a combination of information technology such as the Internet of Things, big data, cloud computing, and artificial intelligence with power technology. It is mainly composed of data acquisition equipment such as sensors, computing equipment such as computers, and physical equipment such as transformers. It deeply integrates advanced technologies such as information collection and transmission, data calculation and processing, and equipment control and decision-making into physical systems such as power-related equipment to realize the interconnection and interaction between people and things in the power system.

[0004] On March 7, 2019, the State Grid Corporation of China released the Outline for the Construction of the Ubiquitous Power Internet of Things, which made work arrangements for the construction and development of the power Internet of Things. According to the document, after three years of hard work, the ubiquitous power Internet of Things will be initially built by 2021; after three years of improvement, the ubiquitous power Internet of Things will be built by 2024. The outline provides the architecture of the power Internet of Things, which is divided into four layers like other forms of Internet of Things systems: perception layer, network layer, platform layer and application layer.

[0005] According to the standards for the architecture and functional requirements of the power Internet of Things, the demand perception of the power Internet of Things should have functions such as data collection, local communication, aggregation and forwarding, centralized verification, edge computing, and data storage. The network layer here pays more attention to the remote communication function of data, and designs the network architecture based on the network-bearing business type, scale, target coverage area, coverage area, capacity, and performance planning goals. The network architecture includes a network architecture based on a private network, a network architecture based on a public network, and a network architecture based on a hybrid network. At the same time, software-defined networks, network virtualization and other technologies can be used to achieve the integration of heterogeneous networks, improve network connection flexibility, improve network resource utilization, and achieve logical isolation of perception layer business data, upload perception layer business data to the platform layer, or transmit platform layer control commands and configuration information to the perception layer.

[0006] The needs that need to be sensed in the power Internet of Things have diverse characteristics. On the power supply side, a large number of new energy power generation equipment such as wind power and photovoltaic power generation are connected, and it is necessary to sense temperature, optics, location and other information, monitor the operating status and health status of the power generation equipment, prevent accidents, improve power generation efficiency and extend equipment life. On the grid side, in scenarios such as power transmission, substation, and distribution, it is necessary to use micro-meteorological, temperature and humidity, tower inclination, icing, sag, wind deviation, partial discharge, vibration and pressure sensing devices to collect grid operation and equipment status, environment and other auxiliary information, and support comprehensive information perception and intelligent application in the production and operation of the grid. On the load side, it is necessary to use sensor measurement devices such as power quality and load monitoring to collect information such as smart electricity consumption and new energy vehicle load, support the utilization of flexible load resources on the demand side, and improve energy utilization and the level of refined management of energy consumption on the user side.

[0007] Current overall solution for power demand sensing

[0008] First, various parameter data in the power system, such as current, voltage, power, etc., are collected through smart meters, sensors or monitoring equipment. After completing pre-processing steps such as data cleaning, denoising, correction and standardization, they are transmitted to the central data storage and processing system through the communication network. In the central data storage and processing system, the data can be organized and managed according to the timestamp to support the query and analysis of historical data. Data analysis technologies such as statistical analysis, machine learning, and time series analysis are applied to perform real-time data processing, load forecasting, anomaly detection, load curve analysis and other operations on the data to understand and extract power demand. According to the data analysis results, based on historical data and model algorithms, the load demand in the future period is predicted. According to the prediction results and constraints, the operation strategy of the power system is adjusted to achieve more efficient, stable and reliable power supply. Finally, based on the demand forecast and optimization results, decision support and control strategies such as load scheduling, power supply strategy, and energy management strategy will be formulated to ensure the normal operation and stability of the power system.

[0009] Routing strategies currently available for power demand sensing services

[0010] Hybrid Routing: It combines the features of static routing and dynamic routing. You can manually configure a portion of static routing, while using dynamic routing protocols to automatically learn and update the routing information of other portions. Static routing associates a specific destination address with an outbound interface, which is suitable for situations where demand preferences do not change frequently, but is not suitable for dynamic network environments; dynamic routing algorithms can dynamically adjust routing paths based on demand preferences, network topology, and link status, which is suitable for situations where demand preferences change frequently and can provide more flexible and adaptive routing forwarding.

[0011] Policy-based Routing: Select a routing path based on a specific policy or demand preference. Traffic can be matched based on conditions such as source IP address, destination IP address, port number, protocol type, etc., and the best routing path can be selected for forwarding based on predefined policies, which can achieve flexible routing selection based on demand preferences.

[0012] SDN routing (Software-Defined Networking Routing): SDN routing is a routing method based on software-defined networking. By separating the network control and data planes, a centralized controller is used to manage and configure routing policies. SDN routing can perform flexible routing forwarding based on demand preferences, network topology, and real-time data.

[0013] Disadvantages of existing technology

[0014] 1. There is a lack of unified description of power grid services and needs, and no consistent specifications and standards have been formed. This will result in high costs and low efficiency in the processing of power grid equipment and systems, and will also pose certain security risks.

[0015] 2. The lack of measurement of the power Internet of Things network status is not conducive to the efficient and reliable transmission of data packets.

[0016] 3. The characteristics of different power demands are not taken into account. The demands in different business scenarios have different requirements for communications and preferences for transmission paths. Failure to consider this condition may cause more data retransmissions and losses.

[0017] 4. It is impossible to carry out sustainable control during the transmission process and to adaptively select routing strategies based on specific business needs and communication requirements, which can easily lead to unbalanced resource utilization, network congestion, and lack of service quality assurance. Summary of the invention

[0018] Design a solution to describe the power grid services and demands, use ENM to measure, and fully obtain the network status information of the power Internet of Things. For the transmission of the power Internet of Things demand-aware services, in order to improve its efficiency and reliability, first, design a new data packet structure, add demand identification and demand attribute information, and then clearly identify and describe the characteristics and requirements of different power grid business needs, and provide more basis for subsequent processing and forwarding; secondly, based on the characteristics of different business needs in the scenarios of generation, transmission, transformation, distribution, and use, make decisions based on the communication requirements of the business needs. Numbering is performed to achieve binding between the two; finally, a demand preference matching table is set in the intermediate device, and the matching relationship between business needs and preferences is obtained according to the matching model, and the demand preference is determined by looking up the table. The best forwarding path is selected in combination with the demand preference to achieve sustainable control and reasonable allocation of resources. The present invention provides a smart collaborative power demand network method.

[0019] The technical solution of the present invention is as follows:

[0020] A smart and coordinated power demand network method, comprising:

[0021] Step (1): According to the needs of power Internet of Things applications, sensors are used to measure various demand indicators in different scenarios such as generation, transmission, transformation, distribution, and use, and the ENM method is used to measure the power grid network status information.

[0022] Step (2): Describe the grid services and requirements, define keywords for each requirement information item, and standardize them, and bind the grid service requirement information and communication requirements through numbering decisions.

[0023] Step (3): Embed the standardized bound demand attribute information into the service request message, encapsulate it into a data packet for transmission, and add corresponding fields in the data packet header to identify and describe the demand information.

[0024] Step (4): When the data packets enter the gateway and intermediate forwarding devices, a sequential caching mechanism is implemented to store the received out-of-order data packets in the cache area and use the sorting logic to determine the correct order of the data packets based on the sequence numbers of the data packets.

[0025] Step (5): Add demand-aware control workflows to gateways and forwarding devices. Data packets are parsed, matched, and de-parsed to perceive the characteristics and requirements of the demand.

[0026] Step (6): Set up a demand preference matching table in the forwarding device, obtain the corresponding relationship between demand and preference through the matching model, and store it in the matching table, and use the EDA information extracted in the previous step to query the table items.

[0027] Step (7): Reassemble the parsed data packets, select the best routing strategy based on demand preferences, and continue transmission according to the requirements of the power grid business needs.

[0028] Technical effects and advantages of the present invention:

[0029] The present invention provides a solution to the power demand processing mechanism based on the intelligent collaboration theory. The ENM method is used to measure the network status information of the power Internet of Things. Based on the characteristics of power demand in different business scenarios, decision numbers are made according to the correspondence between business demand and communication requirements to achieve binding between the two. The matching model is used to match demand with preference to achieve adaptive routing forwarding based on demand preference. Combined with the existing routing strategy, the efficiency of power Internet of Things business transmission is effectively improved, and diversified demand perception in the power Internet of Things can be supported. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of the overall solution of the present invention.

[0031] Figure 2 This is the current structure diagram of the power Internet of Things.

[0032] Figure 3 Schematic diagram of demand perception.

[0033] Figure 4 This is the ENM datagram message format diagram.

[0034] Figure 5 Numbered diagram for business scenarios.

[0035] Figure 6 The following is a schematic diagram of the transmission requirements for specific business needs in the substation scenario.

[0036] Figure 7 Number the decision diagram for the business requirements.

[0037] Figure 8 Schematic diagram of data packet with corresponding fields added.

[0038] Fig. 9 This is a diagram of the demand-aware control workflow. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] Figure 1is a flowchart of an overall solution implementation of a power demand processing mechanism based on intelligent collaboration theory according to an embodiment of the present invention. Figure 2 This is the current structure diagram of the power Internet of Things. Figure 3 This is a schematic diagram of demand perception using this solution. Figure 4 This is the ENM datagram message format diagram. Figure 5 Numbered diagram for business scenarios. Figure 6 This is a schematic diagram of the transmission requirements for specific business needs in the substation scenario. Figure 7 Number the decision diagram for business requirements, Figure 8 To add the corresponding field data packet diagram, Fig. 9 This is a diagram of the demand-aware control workflow. Figures 4 to 9 Yes Figure 3 A concrete supplementary description of .

[0041] like Figure 1 As shown, the present invention provides a solution to the power demand processing mechanism based on the intelligent collaboration theory, which includes seven steps.

[0042] Step (1): A variety of measuring equipment and methods should be adopted to meet different needs. Tunneling magnetoresistance (TMR) magnetic current sensors can be used to detect weak currents, high-precision AC and DC, and electric energy meters in the power grid with high sensitivity. A transformer oil-dissolved gas separation device is used to separate the dissolved gas in the transformer oil using an oil-gas separation membrane to detect the status of the transformer. Through an overhead transmission line dynamic capacity expansion system based on non-contact sensing technology, laser radar and infrared temperature measurement technology are used to obtain key parameters such as the distance between the conductor and the ground and the conductor temperature, so as to evaluate and verify the current carrying capacity of the conductor and provide basic data for transmission line capacity adjustment.

[0043] In addition, the IPv4 network architecture of the existing power communication system faces serious problems such as insufficient address space and poor scalability, and cannot effectively support the ubiquitous access requirements in the power Internet of Things. Compared with IPv4, IPv6 technology has a larger address space and high scalability, as well as optimized network functions and performance. To perceive the state of the power Internet of Things using the IPv6 network architecture, it is first necessary to perform network measurement, timely understand, monitor and master the network status, so as to timely optimize the network structure, improve the network service quality, and realize network fault diagnosis and recovery. Therefore, an Electric Network Measurement (ENM) method is proposed. When a data packet enters the network measurement initiator (Electric Network Measurement Source), through the pre-defined data plane sampling settings, the device will encapsulate an ENM Header in the message, define the ENM structure, add ENM data after the ENM header, and fill the network forwarding device information to be collected into the ENM data structure; then the data packet will be forwarded to the next hop of the path, and according to the data structure defined by ENM, the local ENM metadata information will continue to be added; after the data packet passes through all the devices that need to collect status information, it will be forwarded to the last network device (ENM Sink), and the ENM Sink will remove the ENM Header and send the path information to the monitoring server. In this way, the monitoring server obtains the network link status information, including switch queue depth, queuing delay, link delay and link utilization, etc., which can quickly locate faulty links and discover bottlenecks.

[0044] superior Figure 4 Describes the composition of the ENM datagram, which is mainly divided into 9 parts, 8 bytes of header information and multiple data information, that is, the collected switch device status information. The meanings of the relevant fields of the datagram are as follows:

[0045] 1) The first three fields represent the version number, the copy permission reserved flag, and the number of data types in the data packet, occupying 2 bits, 9 bits, and 5 bits respectively;

[0046] 2) Next is the 8-bit maximum hop count, which indicates the maximum number of datagrams that can be added;

[0047] 3) Next is the 8-bit current total hop count, which indicates the number of datagrams received;

[0048] 4) Datagram content, which indicates the specific content of the measurement, is divided into 8 fields. The first 4 bits represent the switch address, inbound port number, forwarding time, and queue occupancy respectively, and the last 4 bits represent the inbound timestamp, outbound port number, queue congestion status, and outbound port link utilization respectively.

[0049] Step (2): Different demand information is set by defining and standardizing keywords. For example, the keyword "TYPE" is defined and standardized to represent the demand type. The business scenario and demand are represented by numbers to achieve the binding of power grid business demand and transmission requirements. The keyword "SIZE" represents the size of the demand information. The keywords "DATE" and "TIME" represent the effective date and time of the demand content, respectively. The keyword "PRIO" specifies the representation format of the Electricity Demand Attribute (EDA) information as follows:

[0050] EDA={key1:value1,key2:value2,…}

[0051] For example, the sensor measures the voltage in a power transmission scenario. The required content size is 100KB, and the effective date and time is 18:45:23 on January 20, 2024. The business scenarios are numbered in the order of power generation, transmission, substation, distribution, and power consumption. The substation scenario is numbered 3. Taking the substation scenario as an example, Figure 6 As shown, various business needs have different communication requirements.

[0052] go through Figure 7 The decision-making process shown can get the number corresponding to the specific business demand. If the remote signal change alarm business demand in the substation scenario is selected, the demand information can be expressed as:

[0053] EDA

[0054] ={TYPE:35,SIZE:100KB,DATA:20240120,TIME:184523,PRIO:NONE}

[0055] Step (3): Add a demand information field (DI) to the data packet header format. This field is embedded between the data packet header and the actual transmission data information. This field contains an 8-bit binary value, which is used to indicate the length of the information. If the DI field in the data packet is set to 0, it means that the data packet does not contain demand information. Otherwise, it contains demand information of the size indicated by the DI field. According to the original protocol, the added field and EDA content are encapsulated to form a complete data packet.

[0056] Step (4): When data packets enter the gateway and intermediate forwarding devices, a sequential caching mechanism is executed to store the received out-of-order data packets in the cache area and use sorting logic to determine the correct order of the data packets based on the sequence number of the data packets.

[0057] Step (5): Add a demand-aware control workflow to the gateway and forwarding device, which consists of a parser, a matching pipeline, and a de-parser. The function of the parser is to parse the information of the data packet, including MAC address information, IP address information, Port information, and EDA information. The parsed information will be stored as a data structure and type that can be read by the forwarding device. The IP address and Port information are used to update the routing table and select the transmission path. The demand attributes carried by the EDA information will enter the matching pipeline for matching business preferences.

[0058] Step (6): After the data packet is parsed in the previous step, the corresponding relationship between demand and preference needs to be obtained through the matching model. Two sets are defined. One is the demand set D = {d1, d2, ..., d n}, contains n demand elements, preference set P = {p1, p2, ..., p m}, contains m preference elements. Matching matrix C in element C ij Denotes the demand element d i and the preference element p j The degree of match between ij is a decision variable, used to indicate whether to require element d i and the preference element p j Matching is a corresponding relationship, x ij =1 means match, x ij = 0 means no match. The goal is to maximize the sum of the weights of the selected matches in the matching matrix:

[0059]

[0060] The constraints of the matching model are:

[0061] Each requirement element can only be assigned to one preference element, x ij ∈{0,1}

[0062] After matching, the corresponding relationship between the demand and preference is stored in the matching table for subsequent query. After the data packet analysis just now, the key value in the EDA information can be extracted, and the key value is used to query the demand-preference matching table in the forwarding device. If the query is found, the preference of the demand in terms of time, priority, bandwidth, etc. is known, so that the forwarding device will select the path according to the demand preference.

[0063] Step (7): Reassemble the parsed data packets in the order of the protocol and continue to transmit according to the routing forwarding strategy determined in the previous step. When the network changes dynamically, within a single autonomous system (AS), the shortest path tree is calculated using the Dijkstra algorithm based on the network status information in the power grid status database, and the best path is assigned to each target network. Each router notifies each other by exchanging network status update messages. Routing information is exchanged between different autonomous systems (AS) through gateways. Each router maintains a routing table containing network prefixes and path information of other ASs. Path attributes are used to describe AS paths and other attributes to help select the best path. When the network changes dynamically, forwarding rules are defined based on demand preferences. Conditions such as source IP address, destination IP address, protocol, port number, etc. can also be considered to determine which data packets should be routed to a specific path or next hop. On the router or switch, the routing function based on the above rules needs to be configured and associated with a specific path or next hop. When a data packet enters, the router or switch will check the header information of the data packet and match it with the defined rules. If the match is successful, the next hop or path of the data packet is determined according to the established routing configuration. If the match fails, the packet is forwarded according to the default routing table. Flexible adjustment of routing policies can ensure the reasonable allocation of resources.

[0064] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A smart collaborative power demand network method, characterized by: include: Step (1): According to the needs of power Internet of Things applications, sensors are used to measure various demand indicators in different scenarios such as generation, transmission, transformation, distribution, and use, and the ENM method is used to measure the power grid network status information; Step (2): Describe the grid services and requirements, define keywords for each requirement information item, standardize them, and bind the grid service requirement information and communication requirements through numbering decisions; Step (3): embedding the standardized bound demand attribute information into the service request message, encapsulating it into a data packet for transmission, and adding corresponding fields in the data packet header to identify and describe the demand information; Step (4): The data packets enter the gateway and the intermediate forwarding device to execute the sequential caching mechanism, store the received out-of-order data packets in the cache area, and use the sorting logic to determine the correct order of the data packets based on the sequence number of the data packets; Step (5): Add demand-aware control workflows to the gateway and forwarding devices. Data packets are parsed, matched, and de-parsed to perceive the characteristics and requirements of the demand. Step (6): Set a demand preference matching table in the forwarding device, obtain the corresponding relationship between demand and preference through the matching model and store it in the matching table, and use the EDA information extracted in the previous step to query the table item; Step (7): Reassemble the parsed data packets, select the best routing strategy based on demand preferences, and continue transmission according to the requirements of the power grid business needs.

2. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (1) is specifically as follows: a variety of measuring devices and measuring methods should be adopted according to different needs. A tunnel magnetoresistance (TMR) magnetic sensitive current sensor can be used to detect weak currents, high-precision AC and DC, and electric energy meters in the power grid with high sensitivity; a transformer oil-dissolved gas separation device is used to separate the dissolved gas in the transformer oil using an oil-gas separation membrane to detect the status of the transformer; an overhead transmission line dynamic capacity increase system based on non-contact sensing technology is used to obtain key parameters such as the distance between the conductor and the ground and the conductor temperature by relying on laser radar and infrared temperature measurement technology, so as to realize the conductor current carrying capacity evaluation and verification, and provide basic data for transmission line capacity adjustment; In addition, the IPv4 network architecture of the existing power communication system faces serious problems such as insufficient address space and poor scalability, and cannot effectively support the ubiquitous access requirements in the power Internet of Things. Compared with IPv4, IPv6 technology has a larger address space and high scalability, as well as optimized network functions and performance. To perceive the state of the power Internet of Things using the IPv6 network architecture, it is first necessary to perform network measurement to timely understand, monitor and master the network status, so as to timely optimize the network structure, improve the network service quality, and realize network fault diagnosis and recovery. Therefore, an Electric Network Measurement (ENM) method is proposed. When a data packet enters the network measurement initiator (Electric Network Measurement Source), through the pre-defined data plane sampling settings, the device will encapsulate an ENM Header in the message and define the ENM structure at the same time. ENM data is added after the header, and the network forwarding device information that needs to be collected is filled into the ENM data structure; then the data packet will be forwarded to the next hop of the path, and the local ENM metadata information will continue to be added according to the data structure defined by ENM; when the data packet passes through all the devices that need to collect status information, it is forwarded to the last network device (ENM Sink), and ENM Sink removes the ENM Header and sends the path information to the monitoring server. In this way, the monitoring server obtains the network link status information, including switch queue depth, queuing delay, link delay and link utilization, etc., which can quickly locate faulty links and discover bottlenecks.

3. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (2) is specifically as follows: different demand information is set by defining and standardizing keywords. For example, the keyword "TYPE" is defined and standardized to represent the demand type. The business scenario and demand are represented by numbers to achieve the binding of power grid business demand and transmission requirements. The keyword "SIZE" represents the size of the demand information. The keywords "DATE" and "TIME" represent the effective date and time of the demand content respectively. The keyword "PRIO" is specified. The representation format of the electricity demand attribute (EDA) information is as follows: EDA={Key1:value1,Key2:value2,…}.

4. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (3) is specifically as follows: adding a demand information field (DI) to the data packet header format, stipulating that the field is embedded between the data packet header and the actual transmission data information, and the field contains an 8-bit binary value used to indicate the length of the information; if the DI field in the data packet is set to 0, it means that the data packet does not contain demand information, otherwise it contains demand information of the size indicated by the DI field; encapsulating the added field and EDA content according to the original protocol to form a complete data packet.

5. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (5) is specifically as follows: adding a demand-aware control workflow to the gateway and forwarding device, which consists of a parser, a matching pipeline, and a de-parser. The function of the parser is to parse the information of the data packet, including MAC address information, IP address information, Port information, and EDA information. The parsed information will be stored as a data structure and type that can be read by the forwarding device. The IP address and Port information are used to update the routing table and select the transmission path. The demand attributes carried by the EDA information will enter the matching pipeline for matching business preferences.

6. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (6) is as follows: After the data packet is parsed in the previous step, the corresponding relationship between demand and preference needs to be obtained through the matching model. Two sets are defined, one of which is the demand set D = {d1, d2, ..., d n }, contains n demand elements, preference set P = {p1, p2, ..., p m }, contains m preference elements; matching element C in matrix C ij Denotes the demand element d i and the preference element p j The degree of match between ij is a decision variable, used to indicate whether to require element d i and the preference element p j Matching is a corresponding relationship, x ij =1 means match, x ij =0 means no match; The goal is to maximize the sum of the weights of the selected matches in the matching matrix: The constraints of the matching model are: Each requirement element can only be assigned to one preference element, x ij ∈{0,1} After matching, the corresponding relationship between the demand and preference is stored in the matching table for subsequent query. After the data packet analysis just now, the key value in the EDA information can be extracted, and the key value is used to query the demand-preference matching table in the forwarding device. If the query is found, the preference of the demand in terms of time, priority, bandwidth, etc. will be known, so that the forwarding device will select the path according to the demand preference.

7. The method for a smart collaborative power demand network according to claim 1, characterized in that: Step (7) is specifically as follows: reassemble the parsed data packets in the order of the protocol, and continue to transmit according to the routing forwarding strategy determined in the previous step; when the network changes dynamically, within a single autonomous system (AS), the shortest path tree is calculated using the Dijkstra algorithm based on the network status information in the power grid status database, and the best path is assigned to each target network; each router notifies each other by exchanging network status update messages; exchange routing information between different autonomous systems (AS) through gateways; each router maintains a routing table containing network prefixes and path information of other ASs; use path attributes to describe AS paths and other attributes to help select the best path; when the network changes dynamically, define forwarding rules based on demand preferences, and may also consider conditions such as source IP address, destination IP address, protocol, port number, etc. to determine which data packets should be routed to a specific path or next hop; On the router or switch, the routing function based on the above rules needs to be configured and associated with a specific path or next hop; When a data packet enters, the router or switch will check the header information of the data packet and match it with the defined rules; if the match is successful, the next hop or path of the data packet will be determined according to the established routing configuration; if the match is unsuccessful, it will be forwarded according to the default routing table; flexible adjustment of routing policies can ensure the rational allocation of resources.

Images