An adaptive upgrade-based electric energy transaction architecture intelligent optimization system

Abstract

This invention discloses an intelligent optimization system for an adaptively upgraded power trading architecture, comprising a data analysis module and an adaptive optimization module. The adaptive optimization module periodically determines whether to perform adaptive architecture update analysis on a target optimization platform based on update demand coefficients. The target optimization platform has several independent transaction analysis units that perform specific business functions, including but not limited to data acquisition, transaction management, and settlement processing. The update analysis module determines whether to perform container deployment optimization and container execution optimization based on the analysis execution status of each execution container. This invention improves the processing efficiency and analysis quality of multi-source heterogeneous data during the analysis process and makes targeted optimizations to the container deployment scheme of the power trading platform based on actual conditions, ensuring that the transaction decision analysis process can adapt to the dynamic changes of existing market data.

Description

Technical Field

[0001] This invention belongs to the field of electricity trading, specifically relating to an adaptively upgraded intelligent optimization system for electricity trading architecture. Background Technology

[0002] With the deepening of electricity market reform, electricity trading has shown significant characteristics such as diversified participants, diversified trading products, and shorter trading cycles. During the real-time operation of the electricity spot market, high-frequency and large-scale data interaction and decision analysis are required between grid dispatching agencies, power generation companies, electricity sales companies, and electricity users. The data types involved include structured time-series data, such as node voltage and line power in the grid SCADA system; semi-structured quotation format data, such as time-segmented electricity volume and price declarations by market participants; and unstructured policy text data, such as notices of adjustments to trading rules, among other heterogeneous forms.

[0003] To efficiently process the aforementioned multi-source heterogeneous data and support complex trading decisions, existing power trading platforms generally adopt a containerized architecture. This encapsulates independent trading analysis units responsible for specific business functions such as data acquisition, transaction management, and settlement processing within execution containers for unified scheduling and resource management. While containerized deployment offers advantages in improving concurrent processing capabilities and system stability, a fixed container deployment architecture often struggles to adapt effectively to the real-time changes in the actual power trading market environment, such as volatile electricity prices, frequent changes in trading volume, and dynamic policy adjustments. Specifically, when market trading activity surges or data complexity increases abruptly, some execution containers may experience excessive load or intensified resource competition, while others may remain idle. This resource mismatch and architectural lag directly leads to increased data processing latency, decreased analysis efficiency, and even affects the timeliness and accuracy of trading decisions. Therefore, how to construct an optimized system that can adaptively adjust container deployment and execution strategies according to dynamic market changes has become a pressing technical problem to be solved in this field.

[0004] To address the optimization of task processing architecture in power trading platforms, the industry currently employs the following main technical approaches: The first approach combines static resource allocation with manual intervention. Before the platform goes live, this method pre-allocates fixed computing resources to each execution container based on historical data or peak load estimates. When the system detects that the load on some containers consistently exceeds a preset threshold, operations personnel manually trigger scaling operations or adjust container orchestration strategies. This method is simple to implement and is commonly used by small and medium-sized trading platforms.

[0005] The second approach is based on chaincode negotiation and verification technology using distributed databases. For example, some existing solutions employ a distributed ledger architecture, automatically negotiating the electricity volume and price between the buyer and seller through a chaincode model deployed in a blockchain network. It then invokes network security constraint chaincode to perform power flow security verification on temporary contracts, ultimately confirming the verified contracts as valid transactions. This approach utilizes the consensus mechanism of distributed nodes to replace some functions of a centralized scheduling system, thereby improving the automation level and data reliability of the transaction process to a certain extent.

[0006] The third type is an elastic scaling strategy based on preset rules. Some cloud-native trading platforms utilize the horizontal autoscaling mechanism of container orchestration tools such as Kubernetes. By setting thresholds for CPU utilization or memory usage, the number of replicas is automatically increased when the resource usage of a certain execution container exceeds the threshold, and excess replicas are reclaimed when the usage decreases. This method automates the adjustment of the number of containers, reducing the workload of manual operation and maintenance.

[0007] While the aforementioned existing technologies have alleviated the task processing pressure on electricity trading platforms to some extent, they still have the following significant shortcomings when dealing with highly dynamic and complex market environments: First, the response lag of static resource allocation and manual intervention is significant. Electricity market fluctuations are sudden and instantaneous. For example, when extreme weather causes a sharp drop in renewable energy output or a surge in load on the load side, transaction volume and price fluctuations can change drastically within minutes. Manual decision-making and operations cannot accurately adjust the container architecture within such a short time window, easily leading to a backlog of transaction analysis tasks or timeouts in critical data processing, severely impacting the platform's reliability.

[0008] Secondly, while chaincode negotiation technology based on distributed ledgers solves the decentralized trust problem in the transaction process, its focus is on the formation and security verification process of the transaction contract itself, without addressing the dynamic optimization of the underlying task processing architecture. In other words, this solution cannot perceive the actual load status and data processing needs of each execution container, and cannot solve the problems of container resource mismatch and decreased analysis efficiency caused by dynamic changes in market data.

[0009] Furthermore, existing elastic scaling strategies based on preset thresholds suffer from coarse adjustment granularity and a single optimization dimension. These strategies only adjust the number of replicas based on single resource metrics such as CPU or memory, failing to consider processing efficiency bottlenecks caused by data demand conflicts between different transaction analysis units within the execution container, nor do they structurally optimize the container deployment scheme, such as the reallocation and recombination of task units. This results in limited overall processing efficiency improvements in some scenarios, even though the number of containers increases, due to unresolved demand conflicts. Conversely, when the load decreases, simple replica reclamation may mistakenly merge units with high independent execution capabilities, reducing the system's flexibility in responding to subsequent market changes.

[0010] For example, Chinese Patent Publication No. CN112365119A discloses a distributed power trading model for power distribution networks based on a distributed database. The power purchaser and seller send power purchase and sale information through a distributed database network. A power trading negotiation market chaincode model negotiates the transaction volume and price, forming a temporary contract. A network security constraint chaincode model performs security checks on the line power flow and node voltage of the temporary contract. An on-chain / off-chain model determines the security-checked temporary contract as the final contract and executes the transaction. However, the above solution has the following problems: it fails to set targeted architecture deployment optimization methods based on actual market changes and task execution, resulting in poor adaptability of the actual transaction strategy analysis process to real-world application scenarios, and consequently, low data analysis efficiency of the trading platform.

[0011] Therefore, in summary, it is necessary to design an adaptive upgrade-based intelligent optimization system for electricity trading architecture to solve the above problems. Summary of the Invention

[0012] The purpose of this invention is to provide an adaptive upgrade-based intelligent optimization system for the architecture of electricity trading, which overcomes the problem that existing technologies fail to set targeted architecture deployment optimization methods based on actual market changes and task execution, resulting in poor adaptability of the actual trading strategy analysis process to actual application scenarios, and consequently, low data analysis efficiency of the trading platform.

[0013] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: An adaptive upgrade-based intelligent optimization system for power trading architecture includes: The adaptive optimization module is used to calculate the update demand coefficient based on the sum of market dynamic evaluation parameters and data execution parameters within the monitoring period of each target optimization platform at the end of the monitoring period. When the update demand coefficient is greater than the preset update demand coefficient, the monitoring period of the platform is marked as an update optimization period and the architecture update analysis is triggered. The update analysis module, connected to the adaptive optimization module, is used to respond to architecture update analysis and determine whether to perform container deployment optimization or container execution optimization for each execution container based on the analysis execution status of each execution container. The target optimization platform has several independent transaction analysis units for performing specific business functions. The categories of business functions to be performed include data collection, transaction management, and settlement processing, in order to complete the collection and processing of multi-source heterogeneous data involved in the transaction analysis process. The transaction analysis units are encapsulated in execution containers. The analysis execution status is determined based on the reference execution load parameters and the measured execution load difference parameters of each execution container during the update and optimization cycle. The deployment optimization module, connected to the update analysis module, is used to determine the deployment analysis index based on rich parameters of demand characteristics, and to optimize the deployment scheme by updating or elastically expanding the deployment allocation of execution containers in a certain analysis execution state according to the deployment analysis index; the deployment analysis index includes the demand conflict index and the reference demand balance index. The execution optimization module, connected to the update analysis module, is used to elastically shrink execution containers in the second-class analysis execution state based on the reference execution independence index in order to optimize their execution process.

[0014] Preferably, the market dynamic assessment parameter is the ratio of the number of market dynamic data determined within the platform's monitoring period to the total number of different categories of market assessment data used to determine the market dynamic assessment parameter.

[0015] Preferably, for a single market assessment data point, the assessment reference value is based on the absolute value of the difference between its assessment reference value within the monitoring period of this platform and the assessment reference value within the monitoring period of the previous platform. and the assessment reference values ​​within the previous platform's monitoring period. Calculate the dynamic evaluation index: ; The assessment reference value is the average value of the corresponding market assessment data obtained from each time within a platform's monitoring period. It is used to determine the category of market assessment data for market dynamic assessment parameters, including but not limited to: electricity price volatility, trading volume change rate, and the latest release time of electricity trading-related policies. Market assessment data with a dynamic assessment index greater than the preset dynamic assessment index are recorded as market dynamic data; if the assessment reference value in the previous platform monitoring period is zero, the market assessment data is directly determined as market dynamic data.

[0016] Preferably, the data execution parameters are the average of the execution evaluation indices of various operational evaluation data, used to determine the category of operational evaluation data for the data execution parameters, including but not limited to: the average memory load of each transaction analysis unit, data processing latency, transaction matching success rate, and interface call success rate; Calculate the execution evaluation index based on individual operational evaluation data: ; This is the average of the values ​​obtained for each iteration of the operational evaluation data within the analysis period. The execution reference value for this operational evaluation data is the minimum value of the corresponding operational evaluation data under the condition that the operational requirements of the target optimization platform are met.

[0017] Preferably, the update analysis module marks an execution container that meets one of the following conditions as being in a type of analysis execution state: the reference execution load parameter is greater than the preset reference load parameter, or the execution load difference parameter is greater than the preset execution load difference parameter; The update analysis module marks execution containers that simultaneously meet the following conditions as being in a Class II analysis execution state: the reference execution load parameter is less than or equal to the preset reference load parameter, and the execution load difference parameter is less than or equal to the preset execution load difference parameter.

[0018] Preferably, the reference execution load parameter is the average value of the measured execution load parameters obtained at the end of each preset load monitoring cycle within the update and optimization cycle. The execution load difference parameter is the standard deviation of the measured execution load parameters of the execution container in each iteration within the update and optimization cycle. With the average value The ratio: ; At the end of any load monitoring cycle, that cycle is recorded as the optimized execution cycle, and the measured execution load parameters are as follows: The sum of the communication execution load index and the analysis execution load index of the execution container; where the communication execution load index is the maximum percentage of the actual data transmission bandwidth of the execution container relative to the maximum bandwidth allocated to the execution container within the optimized execution cycle, and the analysis execution load index is the maximum CPU utilization of the execution container within the optimized execution cycle.

[0019] Preferably, the deployment optimization module performs container deployment optimization for execution containers in a type of analysis execution state, including: The judgment is based on the demand feature richness parameter of the execution container, where the demand feature richness parameter is the total number of different transaction data categories that each transaction analysis unit within the execution container needs to process; if the demand feature richness parameter is greater than the preset demand feature richness parameter, then the judgment is based on the demand conflict index: If the demand conflict index is greater than the preset demand conflict index, then the deployment allocation of the execution container will be updated. If the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined based on the demand feature enrichment parameters, and the execution container is elastically expanded.

[0020] Preferably, the step of updating the deployment allocation of the execution container by the deployment optimization module includes: Step A1: Extract transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold; Step A2: Migrate one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle; Step A3: Update the resource quota mapping table for each execution container after migration; Among them, the execution container that has completed the deployment and allocation update has a reference execution independence index that is greater than the preset reference execution independence index after the update, and the demand feature richness parameter is less than or equal to the preset demand feature richness parameter.

[0021] Preferably, the demand conflict index is calculated as follows: For a single execution container, the demand conflict index is the degree of communication demand conflict. Conflict degree with analysis requirements sum: Communication demand conflict for: ; ; This represents the percentage of communication requirements within the execution container. The communication execution burden index of the execution container is the proportion of the number of communication demand units corresponding to the execution container in the total number of transaction analysis units corresponding to the execution container. The rated bandwidth for the data transfer process of this execution container. The average of the rated bandwidth for data transfer processes for each execution container. The rated bandwidth is the maximum bandwidth allocated to the corresponding execution container.

[0022] Preferably, analyze the degree of demand conflict. for: ; ; This represents the percentage of analysis requirements within the execution container. The analysis execution burden index is the ratio of the number of analysis demand units corresponding to this execution container to the total number of transaction analysis units corresponding to this execution container. The maximum memory capacity that can be used during data analysis for this execution container. The average of the maximum memory capacity that can be used during data analysis for each execution container.

[0023] Preferably, the communication demand unit is a transaction analysis unit whose transmission delay demand index is greater than a preset transmission delay demand index; the analysis demand unit is a transaction analysis unit whose analysis delay demand index is greater than a preset analysis delay demand index; for a single transaction analysis unit, the processing tasks of the transaction analysis unit include transmission tasks and analysis tasks, the transmission delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each transmission task that needs to be executed by the transaction analysis unit, and the analysis delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each analysis task that needs to be executed by the transaction analysis unit; the maximum delay is determined according to the task completion time limit stipulated in the electricity trading settlement rules.

[0024] Preferably, the deployment optimization module optimizes container deployment for execution containers in a certain analysis execution state, and makes judgments based on the rich parameters of the execution container's requirement characteristics, and further includes: If the richness parameter of demand characteristics is less than or equal to the preset richness parameter of demand characteristics, the judgment is based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The reference demand balance index is the reciprocal of the sum of the communication demand ratio index and the analysis demand ratio index of the execution container.

[0025] Preferably, the container expansion index The calculation method is as follows: ; k represents the resource quota of the execution container before adjustment. The adjusted resource quota for this execution container, which is the maximum memory capacity that the execution container can use during data analysis. When the deployment optimization module has a demand feature richness parameter greater than the preset demand feature richness parameter and the demand conflict index is less than or equal to the preset demand conflict index, The calculation method is as follows: ; The preset expansion coefficient; When the deployment optimization module has a demand feature richness parameter less than or equal to the preset demand feature richness parameter, and a reference demand balance index less than or equal to the preset reference demand balance index, The calculation method is as follows: ; This is the preset expansion coefficient.

[0026] Preferably, the execution optimization module performs container execution optimization for execution containers in the second-class analysis execution state, including: The determination is based on the reference execution independence index of the execution container, which is the average of the execution independence indices of each transaction analysis unit within the execution container. For a single transaction analysis unit, the execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with that transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that the two transaction analysis units need to transmit or process. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined according to the reference execution independence index, and the execution container is elastically shrunk.

[0027] Preferably, the container shrinkage index is: ; For the resource quota before the execution container was adjusted, Adjusted resource quotas for this execution container. satisfy: ; in, This is the preset shrinkage coefficient.

[0028] Furthermore, the transaction analysis unit includes dedicated parsing adapters for structured time-series data, semi-structured quotation format data, and unstructured policy text data, in order to achieve the collection and processing of multi-source heterogeneous data.

[0029] Preferably, a smart optimization method for an adaptively upgraded power trading architecture includes the following steps: S1, Adaptive Optimization: At the end of each preset platform monitoring cycle, the adaptive optimization module calculates the update demand coefficient based on the sum of the market dynamic assessment parameters and data execution parameters within the platform monitoring cycle. When the update demand coefficient is greater than the preset update demand coefficient, the platform monitoring cycle is marked as an update optimization cycle, and architecture update analysis is triggered. S2, Updated Analysis: The update analysis module responds to the architecture update analysis and determines the analysis execution state of each execution container based on the reference execution load parameters and execution load difference parameters of each execution container within the update optimization cycle. Execution containers that satisfy either the reference execution load parameter being greater than a preset reference load parameter or the execution load difference parameter being greater than a preset execution load difference parameter are marked as being in a Class I analysis execution state. Execution containers that satisfy both the reference execution load parameter being less than or equal to the preset reference load parameter and the execution load difference parameter being less than or equal to the preset execution load difference parameter are marked as being in a Class II analysis execution state. S3, Deployment Optimization: The deployment optimization module targets execution containers in a Class I analysis execution state, determines a deployment analysis index based on rich parameters of demand characteristics, and updates or elastically expands the deployment allocation of the execution container according to the deployment analysis index to optimize its deployment scheme; the deployment analysis index includes a demand conflict index and a reference demand balance index; S4, Execution Optimization: The execution optimization module targets execution containers in the second-class analysis execution state and elastically shrinks the execution container according to the reference execution independence index to optimize its execution process.

[0030] Preferably, the deployment optimization step further includes: The judgment is based on the richness parameter of the demand characteristics of the execution container, which is the total number of different transaction data categories that each transaction analysis unit in the execution container needs to process. If the richness parameter of demand features is greater than the preset richness parameter of demand features, the judgment is based on the demand conflict index: if the demand conflict index is greater than the preset demand conflict index, the deployment allocation of the execution container is updated; if the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined according to the richness parameter of demand features, and the execution container is elastically expanded. If the richness parameter of demand features is less than or equal to the preset richness parameter of demand features, the judgment is made based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The deployment allocation update includes: extracting transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold, migrating one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle, and updating the resource quota mapping table of each execution container after migration.

[0031] Preferably, the optimization step further includes: The judgment is based on the reference execution independence index of the execution container. The reference execution independence index is the average of the execution independence indices of each transaction analysis unit in the execution container. The execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with the transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that need to be transmitted or processed by two transaction analysis units. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined based on the reference execution independence index, and the execution container is elastically shrunk.

[0032] Preferably, it also includes a computer device, including a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the aforementioned adaptive upgrade-based intelligent optimization method for power trading architecture.

[0033] Preferably, it further includes a computer-readable storage medium storing computer instructions for causing the computer to execute the adaptive upgrade-based intelligent optimization method for power trading architecture.

[0034] The beneficial effects of the adaptive upgrade-based intelligent optimization system for power trading architecture provided by this invention are as follows: 1. The technical solution of this invention periodically evaluates the update requirements of the task processing architecture of the power trading platform based on the dynamic changes of the actual power trading market and the fluctuation of the acquired data, so as to make timely optimizations to the task processing architecture of the power trading platform, and makes targeted optimizations to the container deployment scheme based on the actual task processing architecture, so that the optimization results of the container deployment scheme are more in line with the actual situation. This invention improves the timeliness and effectiveness of the optimization results of the task processing architecture of the power trading platform.

[0035] 2. The adaptive optimization module of this invention constructs a periodic update demand assessment mechanism, combines market dynamic assessment parameters and data execution parameters to calculate the update demand coefficient, so as to capture market and system operation changes in real time. When the update demand coefficient exceeds the preset value, the update optimization cycle is triggered immediately, avoiding the architecture lag adjustment caused by the traditional fixed architecture. This invention ensures that the processing capacity of each transaction analysis unit in the power trading platform matches the rhythm of market fluctuations, effectively reducing the problem of low data analysis efficiency caused by architecture adaptation lag.

[0036] 3. The update analysis module of this invention divides execution containers into two types of analysis execution states based on reference execution load parameters and execution load difference parameters. It then matches deployment optimization and execution optimization strategies accordingly. For containers with high load or high volatility, the deployment optimization module focuses on demand characteristics and conflicts, and carries out deployment allocation updates or elastic expansion. For containers with low load and low volatility, the execution optimization module performs elastic contraction around the execution independence of the transaction analysis unit. This targeted optimization method avoids resource mismatch caused by uniform optimization, ensuring the matching of container load with the actual situation and avoiding resource waste caused by idle resources. This makes the optimization scheme of each container fit the actual load characteristics.

[0037] 4. In this invention, for execution containers in a certain analysis execution state, the deployment optimization module quantifies the number of data categories that the container needs to process by enriching parameters based on demand characteristics, identifies containers with complex data demands, assesses the degree of demand overlap of transaction analysis units within the container through a demand conflict index, and judges the degree of demand concentration by referring to a demand equilibrium index. This effectively identifies the problems existing in each execution container and determines targeted optimization solutions. This invention effectively resolves the efficiency bottleneck caused by the concentration of communication or analysis demands within containers, improving the effectiveness of the optimization results for the task processing architecture of the power trading platform.

[0038] 5. In this invention, for execution containers in the second-class analysis execution state, the execution optimization module evaluates the degree of overlap in the task processing of units within the container by referencing the execution independence index. The higher the independence index, the lower the inter-unit dependency and the greater the flexibility to respond to market changes. When the reference execution independence index exceeds a preset value, the container shrinkage index is determined according to the positive correlation of the index, so that core processing capabilities are preserved while idle resources are released. This invention avoids the problem of execution containers in the second-class analysis execution state occupying redundant resources despite low load, and improves the effectiveness of the optimization results for the task processing architecture of the power trading platform. Attached Figure Description

[0039] Figure 1 This is a module connection diagram of the adaptive upgrade-based intelligent optimization system for power trading architecture of the present invention. Figure 2 This is a flowchart illustrating the process of determining whether to perform an architecture update analysis for the target optimization platform based on the update requirement coefficient, according to the present invention. Figure 3 This is a flowchart illustrating how the present invention determines and analyzes the execution status based on reference execution load parameters and execution load difference parameters for each execution container. Figure 4 This is a flowchart illustrating the process of determining whether to perform container deployment optimization and container execution optimization based on the analysis and execution status of each execution container in this invention. Figure 5 This is a schematic diagram of the structure of a computer device in an embodiment of the present invention. Detailed Implementation

[0040] Example 1: like Figure 1 As shown, an adaptive upgrade-based intelligent optimization system for power trading architecture includes an adaptive optimization module, an update analysis module, a deployment optimization module, and an execution optimization module.

[0041] The adaptive optimization module is used to calculate the update demand coefficient based on the sum of market dynamic evaluation parameters and data execution parameters within the monitoring period of each target optimization platform at the end of the monitoring period. When the update demand coefficient is greater than the preset update demand coefficient, the monitoring period of the platform is marked as an update optimization period and the architecture update analysis is triggered. like Figure 2 As shown, the update analysis module is connected to the adaptive optimization module, which is used to respond to the architecture update analysis and determine whether to perform container deployment optimization or container execution optimization for each execution container based on the analysis execution status of each execution container. The target optimization platform has several independent transaction analysis units for performing specific business functions. The categories of business functions to be performed include data collection, transaction management, and settlement processing, in order to complete the collection and processing of multi-source heterogeneous data involved in the transaction analysis process. The transaction analysis units are encapsulated in execution containers. The analysis execution status is determined based on the reference execution load parameters and the measured execution load difference parameters of each execution container during the update and optimization cycle. The deployment optimization module, connected to the update analysis module, is used to determine the deployment analysis index based on rich parameters of demand characteristics, and to optimize the deployment scheme by updating or elastically expanding the deployment allocation of execution containers in a certain analysis execution state according to the deployment analysis index; the deployment analysis index includes the demand conflict index and the reference demand balance index. The execution optimization module, connected to the update analysis module, is used to elastically shrink execution containers in the second-class analysis execution state based on the reference execution independence index in order to optimize their execution process.

[0042] Preferably, the market dynamic assessment parameter is the ratio of the number of market dynamic data determined within the platform's monitoring period to the total number of different categories of market assessment data used to determine the market dynamic assessment parameter.

[0043] In this embodiment, a one-cycle platform monitoring cycle is applied. The duration of the platform monitoring cycle can be determined by the user. The higher the user's requirements for the decision analysis efficiency of the target optimization platform, the shorter the duration of the platform monitoring cycle. One platform monitoring cycle duration is provided, which is 2 hours. At the end of each platform monitoring cycle, the update requirement coefficient of the target optimization platform is detected to determine whether to update and optimize the target optimization platform.

[0044] Preferably, for a single market assessment data point, the assessment reference value is based on the absolute value of the difference between its assessment reference value within the monitoring period of this platform and the assessment reference value within the monitoring period of the previous platform. and the assessment reference values ​​within the previous platform's monitoring period. Calculate the dynamic evaluation index: ; The assessment reference value is the average value of the corresponding market assessment data obtained from each time within a platform's monitoring period. It is used to determine the category of market assessment data for market dynamic assessment parameters, including but not limited to: electricity price volatility, trading volume change rate, and the latest release time of electricity trading-related policies. Market assessment data with a dynamic assessment index greater than the preset dynamic assessment index are recorded as market dynamic data; if the assessment reference value in the previous platform monitoring period is zero, the market assessment data is directly determined as market dynamic data.

[0045] Preferably, the data execution parameters are the average of the execution evaluation indices of various operational evaluation data, used to determine the category of operational evaluation data for the data execution parameters, including but not limited to: the average memory load of each transaction analysis unit, data processing latency, transaction matching success rate, and interface call success rate; Calculate the execution evaluation index based on individual operational evaluation data: ; This is the average of the values ​​obtained for each iteration of the operational evaluation data within the analysis period. The execution reference value for this operational evaluation data is the minimum value of the corresponding operational evaluation data under the condition that the operational requirements of the target optimization platform are met.

[0046] Furthermore, the value of the preset update demand coefficient can be determined by the user based on the actual work scenario. For example, the user can set it based on the update analysis records. The higher the user's requirement for the decision analysis efficiency of the target optimization platform, the smaller the value of the preset update demand coefficient. A method for determining the value of the preset update demand coefficient is provided, in which the update analysis records for updating the target optimization platform are recorded as execution reference records, and the minimum value of the update demand coefficient in the execution reference records that meets the user's requirement for the decision analysis efficiency of the target optimization platform is recorded as the preset update demand coefficient. When the system runs for the first time, the preset update demand coefficient is set to an empirical value X, or calculated based on simulated running data.

[0047] like Figure 3 As shown, preferably, the update analysis module marks an execution container that meets one of the following conditions as being in a type of analysis execution state: the reference execution load parameter is greater than the preset reference load parameter, or the execution load difference parameter is greater than the preset execution load difference parameter; The update analysis module marks execution containers that simultaneously meet the following conditions as being in a Class II analysis execution state: the reference execution load parameter is less than or equal to the preset reference load parameter, and the execution load difference parameter is less than or equal to the preset execution load difference parameter.

[0048] In this embodiment, a cyclic load monitoring cycle is applied. The duration of the load monitoring cycle can be determined by the user. The higher the user's requirements for the decision analysis efficiency of the target optimization platform, the shorter the duration of the load monitoring cycle. A load monitoring cycle duration of 15 minutes is provided. At the end of each load monitoring cycle, the execution load parameters of each execution container are detected.

[0049] Preferably, the reference execution load parameter is the average value of the measured execution load parameters obtained at the end of each preset load monitoring cycle within the update and optimization cycle. The execution load difference parameter is the standard deviation of the measured execution load parameters of the execution container in each iteration within the update and optimization cycle. With the average value The ratio: ; At the end of any load monitoring cycle, that cycle is recorded as the optimized execution cycle, and the measured execution load parameters are as follows: The sum of the communication execution load index and the analysis execution load index of the execution container; where the communication execution load index is the maximum percentage of the actual data transmission bandwidth of the execution container relative to the maximum bandwidth allocated to the execution container within the optimized execution cycle, and the analysis execution load index is the maximum CPU utilization of the execution container within the optimized execution cycle.

[0050] Furthermore, the values ​​of the preset reference load parameter and the preset execution load difference parameter can be determined by the user according to the actual working scenario. For example, the user can set them according to the update analysis records. The higher the user's requirements for the decision analysis efficiency of the target optimization platform, the smaller the value of the preset reference load parameter and the smaller the value of the preset execution load difference parameter. A method for determining the value of the preset reference load parameter is provided, in which the maximum value of the reference load parameter of the execution container in the second type of analysis execution state in the update analysis record that satisfies the user's requirements for the decision analysis efficiency of the target optimization platform is recorded as the preset reference load parameter. A method for determining the value of the preset execution load difference parameter is provided, in which the maximum value of the execution load difference parameter of the execution container in the second type of analysis execution state in the update analysis record that satisfies the user's requirements for the decision analysis efficiency of the target optimization platform is recorded as the preset execution load difference parameter.

[0051] like Figure 4 As shown, preferably, the deployment optimization module performs container deployment optimization for execution containers in a type of analysis execution state, including: The judgment is based on the demand feature richness parameter of the execution container, where the demand feature richness parameter is the total number of different transaction data categories that each transaction analysis unit within the execution container needs to process; if the demand feature richness parameter is greater than the preset demand feature richness parameter, then the judgment is based on the demand conflict index: If the demand conflict index is greater than the preset demand conflict index, then the deployment allocation of the execution container will be updated. If the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined based on the demand feature enrichment parameters, and the execution container is elastically expanded.

[0052] Furthermore, for a single execution container, if the execution container is in a certain type of analysis execution state, it indicates that the reference load parameter or the execution load difference parameter of the execution container is large. This indicates that the execution container has a large task processing load in the actual task processing process or that the task processing load fluctuates greatly during the update and optimization cycle. This further indicates that the deployment scheme of the execution container is not well adapted to the current market situation. The richness parameter of the demand feature corresponding to the execution container characterizes the richness of the demand features of the execution container's execution processing tasks, and determines the targeted optimization method accordingly. For a single execution container, the richness parameter of the demand feature is the number of different categories of transaction data that each transaction analysis unit corresponding to the execution container needs to process. In this invention, the categories of transaction data that the transaction analysis unit needs to process include, but are not limited to: the time-sharing declared electricity volume, declared price, time-sharing transaction volume, transaction price, transaction timestamp, active or reactive power of transmission lines, and node voltage of each market participant.

[0053] Preferably, the step of updating the deployment allocation of the execution container by the deployment optimization module includes: Step A1: Extract transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold; Step A2: Migrate one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle; Step A3: Update the resource quota mapping table for each execution container after migration; Among them, the execution container that has completed the deployment and allocation update has a reference execution independence index that is greater than the preset reference execution independence index after the update, and the demand feature richness parameter is less than or equal to the preset demand feature richness parameter.

[0054] Furthermore, for a single execution container, if the richness parameter of the demand features of the execution container is greater than the preset richness parameter of the demand features, it indicates that the types of data involved in the actual task processing of the execution container are relatively rich. This further indicates that there are different communication or data analysis requirements for the execution container in the actual task processing of each transaction analysis unit. Therefore, it is necessary to analyze the degree of overlap of the demand features corresponding to the transaction analysis unit of the execution container, so as to avoid the conflict of demand features in the execution container during the execution of data processing tasks due to the concentration of demand features in the actual processing, thereby affecting the task processing efficiency of the execution container.

[0055] Preferably, the demand conflict index is calculated as follows: For a single execution container, the demand conflict index is the degree of communication demand conflict. Conflict degree with analysis requirements sum: Communication demand conflict for: ; ; This represents the percentage of communication requirements within the execution container. The communication execution burden index of the execution container is the proportion of the number of communication demand units corresponding to the execution container in the total number of transaction analysis units corresponding to the execution container. The rated bandwidth for the data transfer process of this execution container. The average of the rated bandwidth for data transfer processes for each execution container. The rated bandwidth is the maximum bandwidth allocated to the corresponding execution container.

[0056] Preferably, analyze the degree of demand conflict. for: ; ; This represents the percentage of analysis requirements within the execution container. The analysis execution burden index is the ratio of the number of analysis demand units corresponding to this execution container to the total number of transaction analysis units corresponding to this execution container. The maximum memory capacity that can be used during data analysis for this execution container. The average of the maximum memory capacity that can be used during data analysis for each execution container.

[0057] Preferably, the communication demand unit is a transaction analysis unit whose transmission delay demand index is greater than a preset transmission delay demand index; the analysis demand unit is a transaction analysis unit whose analysis delay demand index is greater than a preset analysis delay demand index; for a single transaction analysis unit, the processing tasks of the transaction analysis unit include transmission tasks and analysis tasks, the transmission delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each transmission task that needs to be executed by the transaction analysis unit, and the analysis delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each analysis task that needs to be executed by the transaction analysis unit; the maximum delay is determined according to the task completion time limit stipulated in the electricity trading settlement rules.

[0058] Furthermore, the values ​​of the preset transmission delay requirement index and the preset analysis delay requirement index can be determined by the user according to the actual working scenario. For example, the user can set them according to the update analysis records. The higher the user's requirement for the decision analysis efficiency of the target optimization platform, the smaller the value of the preset transmission delay requirement index and the smaller the value of the preset analysis delay requirement index. One method for setting the preset transmission delay requirement index is to record the minimum value of the transmission delay requirement index of the communication requirement unit in the update analysis record that meets the user's requirement for the decision analysis efficiency of the target optimization platform as the preset transmission delay requirement index. Another method for setting the preset analysis delay requirement index is to record the minimum value of the analysis delay requirement index of the communication requirement unit in the update analysis record that meets the user's requirement for the decision analysis efficiency of the target optimization platform as the preset analysis delay requirement index. Furthermore, the values ​​of the preset demand feature enrichment parameter and the preset demand conflict index can be determined by the user according to the actual work scenario. For example, the user can set them based on the update analysis record. The higher the user's requirement for the decision analysis efficiency of the target optimization platform, the smaller the value of the preset demand conflict index. A method for determining the value of the preset demand feature enrichment parameter is provided, in which the update analysis record that determines whether to perform deployment and allocation updates based on the demand conflict index is recorded as the demand reference record, and the minimum value of the demand feature enrichment parameter in the demand reference record that meets the user's requirement for the decision analysis efficiency of the target optimization platform is recorded as the preset demand feature enrichment parameter. A method for determining the value of the preset demand conflict index is provided, in which the minimum value of the demand conflict index of the execution container that performs deployment and allocation updates in the demand reference record that meets the user's requirement for the decision analysis efficiency of the target optimization platform is recorded as the preset demand conflict index.

[0059] Preferably, the deployment optimization module optimizes container deployment for execution containers in a certain analysis execution state, and makes judgments based on the rich parameters of the execution container's requirement characteristics, and further includes: If the richness parameter of demand characteristics is less than or equal to the preset richness parameter of demand characteristics, the judgment is based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The reference demand balance index is the reciprocal of the sum of the communication demand ratio index and the analysis demand ratio index of the execution container.

[0060] Preferably, the container expansion index The calculation method is as follows: ; k represents the resource quota of the execution container before adjustment. The adjusted resource quota for this execution container, which is the maximum memory capacity that the execution container can use during data analysis. When the deployment optimization module has a demand feature richness parameter greater than the preset demand feature richness parameter and the demand conflict index is less than or equal to the preset demand conflict index, The calculation method is as follows: ; The preset expansion coefficient; When the deployment optimization module has a demand feature richness parameter less than or equal to the preset demand feature richness parameter, and a reference demand balance index less than or equal to the preset reference demand balance index, The calculation method is as follows: ; This is the preset expansion coefficient.

[0061] Furthermore, the container expansion index is positively correlated with the demand feature richness parameter. Specifically, for an execution container whose demand feature richness parameter is greater than the preset demand feature richness parameter, if the demand conflict index of the execution container is less than or equal to the preset demand conflict index, it indicates that although the transaction analysis unit corresponding to the execution container has many different degrees of demand for communication efficiency and analysis efficiency during the recent execution of processing tasks, the degree of conflict between the demands of each transaction analysis unit is low. In this case, the container expansion index is determined only based on the demand feature richness parameter to optimize the task processing capacity of the execution container and avoid the risk of concentrated processing demand caused by subsequent market dynamic changes.

[0062] Furthermore, the container expansion index and the reference demand equilibrium index are negatively correlated.

[0063] Specifically, for a single execution container, if the richness parameter of the required features of the execution container is less than or equal to the preset richness parameter of the required features, it indicates that the transaction analysis unit corresponding to the execution container has relatively few data categories during the execution of the processing task. The adaptability of the execution container to the current stage can be evaluated by analyzing the concentration of the requirements for communication efficiency and analysis efficiency of the transaction analysis unit corresponding to the execution container during the actual execution of the processing task.

[0064] Furthermore, the value of the preset reference demand balance index can be determined by the user based on the actual work scenario. For example, the user can set it based on the update analysis records. The higher the user's requirement for the decision analysis efficiency of the target optimization platform, the smaller the value of the preset reference demand balance index. A method for determining the value of the preset reference demand balance index is provided, in which the update analysis record that determines the container expansion index of the execution container based on the reference demand balance index is recorded as the balance reference record, and the maximum value of the reference demand balance index in the balance reference record that meets the user's requirement for the decision analysis efficiency of the target optimization platform is recorded as the preset reference demand balance index.

[0065] Preferably, the execution optimization module performs container execution optimization for execution containers in the second-class analysis execution state, including: The determination is based on the reference execution independence index of the execution container, which is the average of the execution independence indices of each transaction analysis unit within the execution container. For a single transaction analysis unit, the execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with that transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that the two transaction analysis units need to transmit or process. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined according to the reference execution independence index, and the execution container is elastically shrunk.

[0066] Preferably, the container shrinkage index is: ; For the resource quota before the execution container was adjusted, Adjusted resource quotas for this execution container. satisfy: ; in, This is the preset shrinkage coefficient.

[0067] Furthermore, the container shrinkage index is positively correlated with the reference execution independence index.

[0068] Specifically, for a single execution container, if the execution container is in the second-class analysis execution state, it indicates that the reference load parameter and execution load difference parameter of the execution container are both small. This indicates that the task processing load of the execution container in the actual task processing process is relatively light, and the fluctuation of the task processing load level within the update and optimization cycle is small. In other words, it indicates that the deployment scheme of the execution container is highly adaptable to the data analysis process of the current actual power trading market. Therefore, the degree of crossover in the task processing process between the transaction analysis units corresponding to the execution container is analyzed and characterized based on the reference execution independence index, which represents the execution container's ability to cope with the impact of market dynamic changes in the subsequent stage.

[0069] Furthermore, the value of the preset reference execution independence index can be determined by the user based on the actual work scenario. For example, the user can set it based on the update analysis records. The higher the user's requirements for the decision analysis efficiency of the target optimization platform, the larger the value of the preset reference execution independence index. A method for determining the value of the preset reference execution independence index is provided, in which the update analysis records that perform elastic shrinkage processing on the execution container are recorded as independent reference records, and the minimum value of the reference execution independence index among the independent reference records that meet the user's requirements for the decision analysis efficiency of the target optimization platform is recorded as the preset reference execution independence index.

[0070] Furthermore, the transaction analysis unit includes dedicated parsing adapters for structured time-series data, semi-structured quotation format data, and unstructured policy text data, in order to achieve the collection and processing of multi-source heterogeneous data.

[0071] Preferably, a smart optimization method for an adaptively upgraded power trading architecture includes the following steps: S1, Adaptive Optimization: At the end of each preset platform monitoring cycle, the adaptive optimization module calculates the update demand coefficient based on the sum of the market dynamic assessment parameters and data execution parameters within the platform monitoring cycle. When the update demand coefficient is greater than the preset update demand coefficient, the platform monitoring cycle is marked as an update optimization cycle, and architecture update analysis is triggered. S2, Updated Analysis: The update analysis module responds to the architecture update analysis and determines the analysis execution state of each execution container based on the reference execution load parameters and execution load difference parameters of each execution container within the update optimization cycle. Execution containers that satisfy either the reference execution load parameter being greater than a preset reference load parameter or the execution load difference parameter being greater than a preset execution load difference parameter are marked as being in a Class I analysis execution state. Execution containers that satisfy both the reference execution load parameter being less than or equal to the preset reference load parameter and the execution load difference parameter being less than or equal to the preset execution load difference parameter are marked as being in a Class II analysis execution state. S3, Deployment Optimization: The deployment optimization module targets execution containers in a Class I analysis execution state, determines a deployment analysis index based on rich parameters of demand characteristics, and updates or elastically expands the deployment allocation of the execution container according to the deployment analysis index to optimize its deployment scheme; the deployment analysis index includes a demand conflict index and a reference demand balance index; S4, Execution Optimization: The execution optimization module targets execution containers in the second-class analysis execution state and elastically shrinks the execution container according to the reference execution independence index to optimize its execution process.

[0072] Preferably, the deployment optimization step further includes: The judgment is based on the richness parameter of the demand characteristics of the execution container, which is the total number of different transaction data categories that each transaction analysis unit in the execution container needs to process. If the richness parameter of demand features is greater than the preset richness parameter of demand features, the judgment is based on the demand conflict index: if the demand conflict index is greater than the preset demand conflict index, the deployment allocation of the execution container is updated; if the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined according to the richness parameter of demand features, and the execution container is elastically expanded. If the richness parameter of demand features is less than or equal to the preset richness parameter of demand features, the judgment is made based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The deployment allocation update includes: extracting transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold, migrating one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle, and updating the resource quota mapping table of each execution container after migration.

[0073] Preferably, the optimization step further includes: The judgment is based on the reference execution independence index of the execution container. The reference execution independence index is the average of the execution independence indices of each transaction analysis unit in the execution container. The execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with the transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that need to be transmitted or processed by two transaction analysis units. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined based on the reference execution independence index, and the execution container is elastically shrunk.

[0074] Preferably, it also includes a computer device, including a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the aforementioned adaptive upgrade-based intelligent optimization method for power trading architecture.

[0075] Preferably, it further includes a computer-readable storage medium storing computer instructions for causing the computer to execute the adaptive upgrade-based intelligent optimization method for power trading architecture.

[0076] The system in this embodiment is used to optimize the task execution architecture of an energy trading platform. Through a real-time feedback and dynamic adjustment mechanism, it automatically adjusts the task processing architecture of the trading platform to avoid the failure of the task execution architecture of the energy trading platform to effectively meet the actual market dynamics, resulting in poor analysis efficiency when making actual trading decisions. The energy trading platform with optimized task execution architecture is referred to as the target optimization platform. The target optimization platform has several independent transaction analysis units for performing specific business functions. The categories of business functions to be performed include, but are not limited to, transaction management and settlement processing. Since the target optimization platform has multiple transaction analysis units for performing transaction analysis tasks, containerized deployment can effectively shorten the latency of actual analysis, thereby effectively improving the data analysis efficiency of the actual trading decision-making process. However, using a fixed deployment architecture is prone to blocking of analysis tasks due to insufficient resources or poor adaptability to the current situation, which in turn affects the data analysis efficiency. At least one transaction execution unit is packaged into a container and referred to as an execution container.

[0077] This embodiment uses several update analysis records. Each update analysis record records at least one update requirement coefficient, reference load parameter, execution load difference parameter, requirement feature richness parameter, requirement conflict index, reference requirement balance index, and reference execution independence index during the continuous architecture optimization process for the target optimization platform. Each update analysis record also has a corresponding qualified mark. The qualified mark records whether the decision analysis efficiency of the target optimization platform meets the user's needs. It can be understood that the user can determine whether the decision analysis efficiency of the target optimization platform meets the user's needs based on the self-defined indicators.

[0078] Example 2: like Figure 5 As shown, embodiments of the present invention also provide a computer device. Figure 5 Taking a single processor 10 as an example, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other using different buses and can be mounted on a common motherboard or otherwise installed as needed. The processor can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices, such as display devices coupled to the interface. In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations, for example, as a server array, a group of blade servers, or a multiprocessor system.

[0079] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0080] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0081] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0082] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0083] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0084] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

Claims

1. A smart optimization system for an adaptively upgraded power trading architecture, characterized in that, include: The adaptive optimization module is used to calculate the update demand coefficient based on the sum of market dynamic evaluation parameters and data execution parameters within the monitoring period of each target optimization platform at the end of the monitoring period. When the update demand coefficient is greater than the preset update demand coefficient, the monitoring period of the platform is marked as an update optimization period and the architecture update analysis is triggered. The update analysis module, connected to the adaptive optimization module, is used to respond to architecture update analysis and determine whether to perform container deployment optimization or container execution optimization for each execution container based on the analysis execution status of each execution container. The target optimization platform contains several independent transaction analysis units for performing specific business functions. These transaction analysis units are encapsulated in execution containers. The execution status is determined based on the reference execution load parameters and the measured execution load difference parameters of each execution container during the update and optimization cycle. The deployment optimization module, connected to the update analysis module, is used to determine the deployment analysis index based on rich parameters of demand characteristics, and to optimize the deployment scheme by updating or elastically expanding the deployment allocation of execution containers in a certain analysis execution state according to the deployment analysis index; the deployment analysis index includes the demand conflict index and the reference demand balance index. The execution optimization module, connected to the update analysis module, is used to elastically shrink execution containers in the second-class analysis execution state based on the reference execution independence index in order to optimize their execution process.

2. The energy trading architecture-based intelligent optimization system based on adaptive upgrading as described in claim 1, characterized in that, The market dynamics assessment parameter is the ratio of the number of market dynamics data determined within the platform's monitoring period to the total number of different categories of market assessment data used to determine the market dynamics assessment parameter.

3. The energy trading architecture-based intelligent optimization system according to claim 2, characterized in that, For individual market assessment data, the absolute value of the difference between the assessment reference value within the monitoring period of this platform and the assessment reference value within the monitoring period of the previous platform is used. and the assessment reference values ​​within the previous platform's monitoring period. Calculate the dynamic evaluation index: ； Market assessment data with a dynamic assessment index greater than the preset dynamic assessment index is recorded as market dynamic data. If the assessment reference value in the previous platform monitoring period is zero, the market assessment data is directly determined as market dynamic data.

4. The energy trading architecture-based intelligent optimization system based on adaptive upgrading according to claim 1, characterized in that, The data execution parameters are the average of the execution evaluation indices of various operational evaluation data, used to determine the category of operational evaluation data for the data execution parameters, including but not limited to: the average memory load of each transaction analysis unit, data processing latency, transaction matching success rate, and interface call success rate; Calculate the execution evaluation index based on individual operational evaluation data: ； This is the average of the values ​​obtained for each iteration of the operational evaluation data within the analysis period. The execution reference value for this operational evaluation data is the minimum value of the corresponding operational evaluation data under the condition that the operational requirements of the target optimization platform are met.

5. The energy trading architecture-based intelligent optimization system based on adaptive upgrading as described in claim 1, characterized in that, The update analysis module marks an execution container that meets one of the following conditions as being in a type of analysis execution state: the reference execution load parameter is greater than the preset reference load parameter, or the execution load difference parameter is greater than the preset execution load difference parameter. The update analysis module marks execution containers that simultaneously meet the following conditions as being in a Class II analysis execution state: the reference execution load parameter is less than or equal to the preset reference load parameter, and the execution load difference parameter is less than or equal to the preset execution load difference parameter.

6. The energy trading architecture-based intelligent optimization system according to claim 5, characterized in that, The reference execution load parameter is the average value of the measured execution load parameters obtained at the end of each preset load monitoring cycle within the update and optimization cycle. The execution load difference parameter is the standard deviation of the measured execution load parameters of the execution container in each iteration within the update and optimization cycle. With the average value The ratio: ； At the end of any load monitoring cycle, that load monitoring cycle is recorded as the optimized execution cycle.

7. The energy trading architecture-based intelligent optimization system according to claim 1, characterized in that, The deployment optimization module optimizes container deployment for execution containers in a certain analysis execution state, including: The judgment is based on the demand feature richness parameter of the execution container, where the demand feature richness parameter is the total number of different transaction data categories that each transaction analysis unit within the execution container needs to process; if the demand feature richness parameter is greater than the preset demand feature richness parameter, then the judgment is based on the demand conflict index: If the demand conflict index is greater than the preset demand conflict index, then the deployment allocation of the execution container will be updated. If the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined based on the demand feature enrichment parameters, and the execution container is elastically expanded.

8. The energy trading architecture-based intelligent optimization system according to claim 7, characterized in that, The steps for updating the deployment allocation of execution containers by the deployment optimization module include: Step A1: Extract transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold; Step A2: Migrate one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle; Step A3: Update the resource quota mapping table for each execution container after migration; Among them, the execution container that has completed the deployment and allocation update has a reference execution independence index that is greater than the preset reference execution independence index after the update, and the demand feature richness parameter is less than or equal to the preset demand feature richness parameter.

9. A smart optimization system for an adaptively upgraded power trading architecture according to claim 7, characterized in that, The method for calculating the demand conflict index is as follows: For a single execution container, the demand conflict index is the degree of communication demand conflict. Conflict degree with analysis requirements sum: Communication demand conflict for: ； ； This represents the percentage of communication requirements within the execution container. The communication execution burden index of the execution container is the proportion of the number of communication demand units corresponding to the execution container in the total number of transaction analysis units corresponding to the execution container. The rated bandwidth for the data transfer process of this execution container. The average of the rated bandwidth for data transfer processes for each execution container. The rated bandwidth is the maximum bandwidth allocated to the corresponding execution container.

10. A smart optimization system for an adaptively upgraded power trading architecture according to claim 9, characterized in that, Analyze the degree of demand conflict for: ； ； This represents the percentage of analysis requirements within the execution container. The analysis execution burden index is the ratio of the number of analysis demand units corresponding to this execution container to the total number of transaction analysis units corresponding to this execution container. The maximum memory capacity that can be used during data analysis for this execution container. The average of the maximum memory capacity that can be used during data analysis for each execution container.

11. The energy trading architecture-based intelligent optimization system according to claim 9, characterized in that, The communication demand unit is a transaction analysis unit whose transmission delay demand index is greater than the preset transmission delay demand index; the analysis demand unit is a transaction analysis unit whose analysis delay demand index is greater than the preset analysis delay demand index; for a single transaction analysis unit, the processing tasks of the transaction analysis unit include transmission tasks and analysis tasks. The transmission delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each transmission task that needs to be executed by the transaction analysis unit, and the analysis delay demand index is the reciprocal of the minimum value among the maximum delays corresponding to each analysis task that needs to be executed by the transaction analysis unit; the maximum delay is determined according to the task completion time limit stipulated in the electricity trading settlement rules.

12. The energy trading architecture-based intelligent optimization system according to claim 10, characterized in that, The deployment optimization module optimizes container deployment for execution containers in a certain analysis execution state, and makes judgments based on the rich parameters of the execution container's requirement characteristics. It also includes: If the richness parameter of demand characteristics is less than or equal to the preset richness parameter of demand characteristics, the judgment is based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The reference demand balance index is the reciprocal of the sum of the communication demand ratio index and the analysis demand ratio index of the execution container.

13. The energy trading architecture-based intelligent optimization system according to claim 12, characterized in that, Container Expansion Index The calculation method is as follows: ； k represents the resource quota of the execution container before adjustment. The adjusted resource quota for this execution container, which is the maximum memory capacity that the execution container can use during data analysis. When the deployment optimization module has a requirement feature richness parameter greater than the preset requirement feature richness parameter and the requirement conflict index is less than or equal to the preset requirement conflict index, The calculation method is as follows: ； This is the preset expansion coefficient; When the deployment optimization module has a demand feature richness parameter less than or equal to the preset demand feature richness parameter, and a reference demand balance index less than or equal to the preset reference demand balance index, The calculation method is as follows: ； This is the preset expansion coefficient.

14. The energy trading architecture-based intelligent optimization system according to claim 13, characterized in that, The execution optimization module performs container execution optimization for execution containers in the second-class analysis execution state, including: The determination is based on the reference execution independence index of the execution container, which is the average of the execution independence indices of each transaction analysis unit within the execution container. For a single transaction analysis unit, the execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with that transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that the two transaction analysis units need to transmit or process. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined according to the reference execution independence index, and the execution container is elastically shrunk.

15. A smart optimization system for an adaptively upgraded power trading architecture according to claim 14, characterized in that, The container shrinkage index is: ； For the resource quota before the execution container was adjusted, Adjusted resource quotas for this execution container. satisfy: ； in, This is the preset shrinkage coefficient.

16. A smart optimization method for an adaptively upgraded power trading architecture according to any one of claims 1-15, characterized in that, Includes the following steps: S1, Adaptive Optimization: At the end of each preset platform monitoring cycle, the adaptive optimization module calculates the update demand coefficient based on the sum of the market dynamic assessment parameters and data execution parameters within the platform monitoring cycle. When the update demand coefficient is greater than the preset update demand coefficient, the platform monitoring cycle is marked as an update optimization cycle, and architecture update analysis is triggered. S2, Updated Analysis: The update analysis module responds to the architecture update analysis and determines the analysis execution state of each execution container based on the reference execution load parameters and execution load difference parameters of each execution container within the update optimization cycle. Execution containers that satisfy either the reference execution load parameter being greater than a preset reference load parameter or the execution load difference parameter being greater than a preset execution load difference parameter are marked as being in a Class I analysis execution state. Execution containers that satisfy both the reference execution load parameter being less than or equal to the preset reference load parameter and the execution load difference parameter being less than or equal to the preset execution load difference parameter are marked as being in a Class II analysis execution state. S3, Deployment Optimization: The deployment optimization module targets execution containers in a Class I analysis execution state, determines a deployment analysis index based on rich parameters of demand characteristics, and updates or elastically expands the deployment allocation of the execution container according to the deployment analysis index to optimize its deployment scheme; the deployment analysis index includes a demand conflict index and a reference demand balance index; S4, Execution Optimization: The execution optimization module targets execution containers in the second-class analysis execution state and elastically shrinks the execution container according to the reference execution independence index to optimize its execution process.

17. The intelligent optimization method for an adaptively upgraded power trading architecture according to claim 16, characterized in that, The deployment optimization steps further include: The judgment is based on the richness parameter of the demand characteristics of the execution container, which is the total number of different transaction data categories that each transaction analysis unit in the execution container needs to process. If the richness parameter of demand features is greater than the preset richness parameter of demand features, the judgment is based on the demand conflict index: if the demand conflict index is greater than the preset demand conflict index, the deployment allocation of the execution container is updated; if the demand conflict index is less than or equal to the preset demand conflict index, the container expansion index is determined according to the richness parameter of demand features, and the execution container is elastically expanded. If the richness parameter of demand features is less than or equal to the preset richness parameter of demand features, the judgment is made based on the reference demand balance index. If the reference demand balance index is less than or equal to the preset reference demand balance index, the container expansion index is determined according to the reference demand balance index, and the execution container is elastically expanded. The deployment allocation update includes: extracting transaction analysis unit pairs within the execution container whose overlap of demand characteristics exceeds a preset overlap threshold, migrating one of the transaction analysis units in the transaction analysis unit pair to a newly created execution container or an existing execution container with the smallest reference execution load parameter during the update and optimization cycle, and updating the resource quota mapping table of each execution container after migration.

18. A smart optimization method for an adaptively upgraded power trading architecture according to claim 16, characterized in that, The optimization steps further include: The judgment is based on the reference execution independence index of the execution container. The reference execution independence index is the average of the execution independence indices of each transaction analysis unit in the execution container. The execution independence index is the reciprocal of the number of other transaction analysis units that have data interaction with the transaction analysis unit. Data interaction refers to the overlap of the types of transaction data that need to be transmitted or processed by two transaction analysis units. If the reference execution independence index is greater than the preset reference execution independence index, then the container shrinkage index is determined based on the reference execution independence index, and the execution container is elastically shrunk.

19. A computer device, characterized in that, It includes a memory and a processor, which are interconnected and communicate with each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the adaptive upgrade-based intelligent optimization method for power trading architecture as described in any one of claims 16 to 18.

20. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the adaptive upgrade-based smart optimization method for power trading architecture as described in any one of claims 16 to 18.

Images