Regional energy management optimization method based on edge computing and multi-objective optimization

Abstract

The present invention belongs to the field of energy management technology, and specifically relates to a regional energy management optimization method based on edge computing and multi-objective optimization, the steps of which include: determining edge devices with qualified processor performance based on processor performance parameters of edge devices, establishing edge device location maps based on the geographical locations of qualified edge devices, applying the initial multi-objective optimization scheme to a regional energy simulation model, obtaining energy efficiency evaluation values ​​and energy efficiency comprehensive difference values, comparing source efficiency comprehensive difference values ​​with energy efficiency comprehensive difference standard values ​​stored in a database, matching pre-selected management optimization schemes and applying them to a regional energy simulation model, and adjusting optimization scheme decisions based on the optimal management optimization scheme according to real-time evaluation results. The present invention can improve the flexibility of overall regional energy management, and is conducive to maintaining the sustainable development of long-term regional energy management.

Description

Technical Field

[0001] The present invention belongs to the field of energy management technology, and specifically relates to a regional energy management optimization method based on edge computing and multi-objective optimization. Background Art

[0002] Regional energy management is the systematic planning, monitoring and optimization of energy production, distribution, consumption and related activities. Through the reasonable planning and management of energy resources in the region, abnormal energy management conditions can be discovered in time, energy management equipment failure or damage can be prevented, management and maintenance costs can be reduced, regional energy management efficiency can be improved, the use of renewable energy can be promoted, energy utilization efficiency can be improved, and resource waste can be reduced to achieve efficient energy utilization, environmental protection and economic sustainable development.

[0003] However, the existing regional energy management methods still have some shortcomings: in regional energy management, edge devices only provide data transmission and computing functions, which are separated from the actual regional energy management decision-making, resulting in delays in the transmission of real-time data, affecting the timeliness of decision-making, and even leading to erroneous energy management decisions. This not only weakens the important influence of edge device processing capabilities in regional energy management, but also wastes computing resources at the decision-making level and fails to realize the potential added value of edge computing. Moreover, the non-participation of edge devices in the decision-making process will lead to reliance on centralized systems at the management level, which may reduce decision-making efficiency, and fail to respond quickly to emergencies in complex environments, making it difficult for optimization plans to meet energy needs. Summary of the invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a regional energy management optimization method based on edge computing and multi-objective optimization, which helps to quickly respond to regional energy environment changes and energy management decision-making needs, and more accurately evaluate the consumption of regional energy resources, thereby achieving more optimized resource allocation and utilization.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] The regional energy management optimization method based on edge computing and multi-objective optimization includes the following steps:

[0007] The regional energy data collected within the regional energy management scope is sent from the data processing center to each edge device, the processor performance parameters of any edge device are obtained, and the edge devices with qualified processor performance are determined based on the processor performance parameters, recorded as qualified edge devices, and the edge devices with unqualified processor performance are replaced;

[0008] Obtaining initial processor qualification levels for qualified edge devices;

[0009] Establish an edge device location map based on the geographic locations of the qualified edge devices, and mark the initial processor qualification level of each qualified edge device on the edge device location map;

[0010] Obtaining a regional energy simulation model stored in a database, applying an initial multi-objective optimization scheme to the regional energy simulation model, obtaining an energy efficiency evaluation value of the regional energy simulation model, and obtaining an energy efficiency comprehensive difference value based on the energy efficiency evaluation value;

[0011] If the energy efficiency comprehensive difference value is not greater than the energy efficiency comprehensive difference standard value stored in the database, the initial multi-objective optimization scheme is applied to the actual regional energy;

[0012] If the energy efficiency comprehensive difference value is greater than the energy efficiency comprehensive difference standard value stored in the database, multiple pre-selected management optimization schemes corresponding to the initial multi-objective optimization scheme stored in the database are matched by the energy efficiency comprehensive difference value, and applied to the regional energy simulation model one by one;

[0013] Obtain the optimized comprehensive difference value of energy efficiency after each pre-selected management optimization scheme is applied to the regional energy simulation model and the processor performance index of each edge device in the device site map;

[0014] The optimal management optimization scheme is determined from the pre-selected management optimization schemes based on the comprehensive difference value of energy efficiency after optimization and the processor performance index of each edge device in the device site map.

[0015] Preferably, the edge device with qualified processor performance is determined based on the processor performance parameters, and the specific analysis process is as follows:

[0016] Obtain the processor performance parameters of any edge device within the regional energy management range, including clock frequency deviation value, thread average response rate, and instruction hit rate;

[0017] Based on the processor performance parameters, the compensation factor parameters set by the edge device in the database are obtained, where the compensation factor parameters include a clock frequency deviation value compensation factor, a thread average response rate compensation factor, and an instruction hit rate compensation factor;

[0018] Obtaining a processor performance index of the edge device based on the processor performance parameter and the compensation factor parameter;

[0019] Extract the edge device processor performance standard value stored in the database, and compare the processor performance index with the edge device processor performance standard value:

[0020] If the processor performance index of the edge device is less than the edge device processor performance standard value, the edge device processor is marked as unqualified;

[0021] If the processor performance index of the edge device is not less than the standard value of the edge device processor performance, the edge device processor is marked as qualified and determined to be an edge device with qualified processor performance.

[0022] Preferably, the calculation formula of the processor performance index of the edge device is as follows:

[0023] ;

[0024] In the formula, α is the performance index of the edge device processor, p is the clock frequency deviation value, s is the average thread response rate, m is the instruction hit rate, and k is the average thread response rate. p is the clock frequency deviation compensation factor set in the database, k s k is the average response rate compensation factor of threads set in the database, m is the instruction hit rate compensation factor set in the database, and e is a natural constant.

[0025] Preferably, the obtaining of the initial processor qualification level of the qualified edge device includes the following specific analysis process:

[0026] Calculate the processor performance deviation rate based on the processor performance index of the qualified edge device and the edge device processor performance standard value stored in the database;

[0027] Extracting the processor performance deviation rate grading interval endpoints stored in the database, the processor performance deviation rate grading interval endpoints stored in the database are the basis for grading the processor qualification level;

[0028] The processor qualification level is graded based on the processor performance deviation rate and the processor performance deviation rate grading interval endpoints stored in the database to obtain an initial processor qualification level.

[0029] Preferably, the energy efficiency evaluation value of the regional energy simulation model is obtained, and the energy efficiency comprehensive difference value is obtained based on the energy efficiency evaluation value of the regional energy simulation model. The specific analysis process is:

[0030] Obtaining an energy efficiency data set after the initial multi-objective optimization scheme is applied to the regional energy simulation model;

[0031] Obtaining a correction factor data set set in a database;

[0032] Obtaining an energy efficiency evaluation value based on an energy efficiency data set and a correction factor data set set in a database;

[0033] Obtaining energy efficiency target values ​​stored in a database;

[0034] The energy efficiency comprehensive difference value is calculated based on the energy efficiency evaluation value, the energy efficiency target value and the energy efficiency difference evaluation model.

[0035] Preferably, the energy efficiency data set specifically includes power generation and energy storage ratio, energy stability range, energy load rate, unit carbon emission intensity, and renewable energy ratio;

[0036] The correction factor data set includes the power generation and energy storage ratio correction factor, energy stability extreme correction factor, energy load rate correction factor, unit carbon emission intensity correction factor, and renewable energy ratio correction factor.

[0037] Preferably, the specific process of obtaining the energy load rate is:

[0038] Obtain energy load parameters after the initial multi-objective optimization scheme is applied to the regional energy simulation model, the energy load parameters including power peak-valley load difference, power generation capacity factor, grid frequency fluctuation rate, and power supply redundancy level;

[0039] The energy load parameter is stored as an energy load designated tag, and the energy load designated tag is compared with the energy load rate corresponding to each energy load designated tag stored in the database to obtain the energy load rate corresponding to the energy load designated tag.

[0040] Preferably, the calculation formula of the energy efficiency evaluation value is:

[0041] ;

[0042] In the formula, is the energy efficiency evaluation value, k C k is the power generation energy storage ratio correction factor, J is the energy stability extreme correction factor, k F is the energy load rate correction factor, k Q is the unit carbon emission intensity correction factor, k Z is the correction factor for the proportion of renewable energy, C is the proportion of power generation and energy storage, J is the energy stability range, F is the energy load rate, Q is the unit carbon emission intensity, Z is the proportion of renewable energy, and e is a natural constant.

[0043] Preferably, the method of determining the optimal management optimization scheme from pre-selected management optimization schemes based on the optimized comprehensive difference value of energy efficiency and the processor performance index of each edge device in the device site map comprises the following steps:

[0044] Determine the edge device qualification rate based on the processor performance index of each edge device;

[0045] Obtain the processor qualification level of the qualified edge device;

[0046] Compare and analyze the processor qualification level of the qualified edge device with the initial processor qualification level to determine the qualification level change coefficient of the qualified edge device;

[0047] Comprehensively analyze the comprehensive difference value of energy efficiency after optimization, the qualified rate of edge devices, and the qualified grade variation coefficient of qualified edge devices to obtain a comprehensive evaluation coefficient;

[0048] The pre-selected management optimization plan corresponding to the smallest comprehensive evaluation coefficient is determined as the optimal management optimization plan.

[0049] Preferably, the calculation formula of the comprehensive evaluation coefficient is:

[0050] ;

[0051] In the formula, is the comprehensive evaluation coefficient, CY is the comprehensive difference value of energy efficiency after optimization, is the edge device qualification rate, is the coefficient of variation of the qualified grade of qualified edge devices, and e is a natural constant.

[0052] The present invention has the following beneficial effects:

[0053] The present invention can quantify the processor performance indicators of edge devices, make qualified judgments on edge devices, and replace unqualified edge devices. The processor performance of edge devices directly affects the processing efficiency and data processing capabilities of edge devices for regional energy resource consumption assessment and management decision-making. Edge devices with qualified processors have lower data transmission and processing delays, and can quickly process large amounts of data, thereby shortening the time for data analysis and decision-making, which is crucial for real-time monitoring and response systems, and helps to quickly respond to regional energy environment changes and energy management decision-making needs, and more accurately evaluate the consumption of regional energy resources, thereby achieving more optimized resource allocation and utilization, and improving overall energy efficiency. At the same time, it can quickly judge the scalability of edge devices with qualified processors, quickly predict the number of sensors and peripheral access that can be supported, adapt to changing energy management needs and monitoring technology advances, so as to maintain the sustainable development of long-term regional energy management.

[0054] The present invention utilizes the pre-selected management optimization schemes stored in the database, combines them with the comprehensive difference values ​​of energy efficiency for intelligent matching, and establishes a dynamic feedback mechanism. This mechanism allows the system to continuously iterate and improve the optimization scheme to adapt to changes in the environment and demand, making the optimization process more accurate and efficient. By gradually applying and evaluating each pre-selected management optimization scheme, a hierarchical and systematic optimization strategy is formed under the unified standard of quantitative data evaluation, which can accurately find the best solution and avoid the risk of blind optimization. It can significantly optimize the overall regional energy consumption and improve energy utilization efficiency.

[0055] When determining the optimal management optimization plan, the present invention not only considers the comprehensive difference value of energy efficiency, but also incorporates the processor performance index of the edge device into the evaluation, so that the optimization plan not only performs well in terms of energy consumption, but also ensures the operating efficiency of the overall energy management. By combining the simulation model and the edge computing framework for rapid evaluation calculation and accurate verification, the risks that may arise in actual applications can be effectively reduced, the reliability and effectiveness of the final plan can be ensured, the optimization plan decision can be adjusted according to the real-time evaluation results, and the flexibility of the overall regional energy management can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0057] Figure 2 Diagram of steps to obtain an initial processor qualification level for a qualified edge device;

[0058] Figure 3 The graph shows the edge device processor performance index and clock frequency deviation value changes. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings.

[0060] like Figure 1As shown in FIG, the regional energy management optimization method based on edge computing and multi-objective optimization includes: the regional energy data collected within the scope of regional energy management is sent from the data processing center to each edge device, and each energy device (such as solar panels, wind turbines, energy storage equipment, etc.) and sensor in the region continuously collect energy consumption, output and environmental data, which are initially sent to the regional data acquisition system through the local area network or directly uploaded to the data processing center. The data processing center cleans, processes and analyzes the data, including noise removal, data standardization, anomaly detection and trend analysis. The processed data and control commands are sent back to the edge devices in the region through a secure communication protocol (such as MQTT, CoAP or HTTPS). The edge devices usually have a certain computing power and can respond to the commands issued by the center in real time, perform local control and optimization, and reduce the load of the data center. Such edge computing can improve the response speed and reliability of the system because it reduces the dependence on the central server and reduces communication delay.

[0061] When the edge device performs related calculations, the processor performance parameters of any edge device are obtained, and the edge devices with qualified processor performance are determined based on the processor performance parameters and recorded as qualified edge devices, and the edge devices with unqualified processor performance are replaced.

[0062] The specific analysis process is as follows: obtain the processor performance parameters of any edge device within the regional energy management range. The processor performance parameters include clock frequency deviation value, thread average response rate, and instruction hit rate. Among them, the clock frequency deviation value refers to the difference between the actual clock frequency of the processor and the nominal clock frequency. The calculation process of the clock frequency deviation value is to first obtain the absolute value of the difference between the actual clock frequency and the nominal clock frequency, and then perform a ratio operation on the absolute value of the difference and the nominal clock frequency to obtain the clock frequency deviation value. The thread average response rate refers to the average response speed of the processor to thread requests within a specific processor operation cycle, which is usually measured in terms of the number of requests processed per second. The thread average response rate is an important indicator for measuring the multi-threaded processing capability of the processor. The instruction hit rate refers to the proportion of the processor that can successfully find the required instructions from the cache during execution. Its value is equal to the ratio of the number of cache hits to the total number of instruction accesses. A high instruction hit rate means that there is only a small delay in the process of successfully finding the required instruction, which directly affects the speed at which the processor executes instructions by reflecting the efficiency of the cache.

[0063] Based on the processor performance parameters, the compensation factor parameters set by the edge device in the database are obtained, and the compensation factor parameters include the clock frequency deviation value compensation factor, the thread average response rate compensation factor, and the instruction hit rate compensation factor. According to the historical data, a mapping set of the historically measured clock frequency deviation value, thread average response rate, instruction hit rate and clock frequency deviation value compensation factor, thread average response rate compensation factor, and instruction hit rate compensation factor is established to obtain the compensation factors corresponding to the current clock frequency deviation value, thread average response rate, and instruction hit rate. The remaining compensation factors in this embodiment are obtained in the same way.

[0064] The processor performance index of the edge device is obtained based on the processor performance parameters and the compensation factor parameters; the edge device processor performance standard value stored in the database is extracted, and the processor performance index is compared with the edge device processor performance standard value: if the edge device processor performance index is less than the edge device processor performance standard value, the edge device processor is marked as unqualified; if the edge device processor performance index is not less than the edge device processor performance standard value, the edge device processor is marked as qualified and determined to be an edge device with qualified processor performance.

[0065] Incorporating compensation factors into the processor performance index calculation can better consider other influencing factors such as environmental changes, making processor performance more comprehensive and detailed; extracting the stored edge device processor performance standard values ​​from the database realizes data-driven performance comparison, can analyze and make decisions based on historical performance data, can dynamically reflect the actual performance of edge devices, and improve the scientific nature of the evaluation; adopting a qualified and unqualified marking mechanism, it realizes the classification management of edge device performance, facilitates further energy resource allocation and scheduling optimization, realizes more accurate energy resource scheduling and task allocation, ensures that high-performance equipment bears heavier loads in key energy management tasks, and improves energy management efficiency; through the marking of unqualified equipment, it can timely discover equipment with insufficient performance, so that it can be maintained or replaced, and improve the reliability of the overall system.

[0066] The calculation formula of the processor performance index of the edge device is as follows:

[0067] ;

[0068] In the formula, α is the performance index of the edge device processor, p is the clock frequency deviation value, s is the average thread response rate, m is the instruction hit rate, and k is the average thread response rate. p is the clock frequency deviation compensation factor set in the database, k s k is the average response rate compensation factor of threads set in the database, m is the instruction hit rate compensation factor set in the database, and e is a natural constant.

[0069] The higher the clock frequency, the faster the processor executes instructions. The larger p is, the greater the degradation of the actual processor performance is, and the fewer instructions the processor can execute per unit time. The increase in p causes the processor to become unstable or slower when obtaining instructions, which will affect the effective reading and execution of instructions, and thus cause m to decrease. s reflects the multitasking capability of the processor, and m refers to the ratio of instructions that hit the cache. A higher m means that most instructions are found in the processor cache, thereby speeding up the execution process and making the processor more powerful. Increasing m can reduce memory access latency, reduce waiting time, and improve thread processing efficiency, thereby improving s. At the same time, the larger p is, the instability of the processor operation will also affect s, resulting in higher latency.

[0070] Figure 3 The graph shows the edge device processor performance index as the clock frequency deviation value changes, where the x-axis represents the clock frequency deviation value and the y-axis represents the edge device processor performance index, which can help intuitively understand how the clock frequency deviation value affects the edge device processor performance index. The larger the clock frequency deviation value, the smaller the edge device processor performance index. As the clock frequency deviation value increases, the impact of the clock frequency deviation value on the edge device processor performance index gradually weakens. The average thread response rate is set to 10, the instruction hit rate is set to 0.8, the clock frequency deviation value compensation factor set in the database is set to 0.6, the average thread response rate compensation factor set in the database is set to 0.8, and the instruction hit rate compensation factor set in the database is set to 0.7. Only the size of the clock frequency deviation value is changed. The example values ​​of the edge device processor performance index are as follows (n is the data group number):

[0071] Table 1 Example values ​​of clock frequency deviation in edge device processor performance index

[0072]

[0073] like Figure 2 As shown, the initial processor qualification level of the qualified edge device is obtained. The specific analysis process is: the processor performance deviation rate is calculated based on the processor performance index of the qualified edge device and the edge device processor performance standard value stored in the database. The calculation process of the processor performance deviation rate is: first obtain the absolute value of the difference between the processor performance index of the qualified edge device and the edge device processor performance standard value stored in the database, and perform a ratio operation on the absolute value of the difference with the edge device processor performance standard value stored in the database to obtain the processor performance deviation rate, thereby quantifying the difference between the actual performance of the qualified edge device processor and the standard performance.

[0074] Extract the processor performance deviation rate grading interval endpoints stored in the database, which are the basis for grading the processor qualification grade; perform processor qualification grade grading based on the processor performance deviation rate and the processor performance deviation rate grading interval endpoints stored in the database to obtain an initial processor qualification grade.

[0075] Extract information on the endpoints of the processor performance deviation rate grading interval from the database, where each interval may include a lower limit and an upper limit, for example, interval 1: lower limit 1.0%, upper limit 5.0%, interval 2: lower limit 5.1%, upper limit 10.0%, interval 3: lower limit 10.1%, upper limit 20.0%; compare the obtained processor performance deviation rate with the extracted interval endpoints, if the performance deviation rate is within the range of interval 1, the qualified level is set to 1, if the performance deviation rate is within the range of interval 2, the qualified level is set to 2, if the performance deviation rate is within the range of interval 3, the qualified level is set to 3, and so on, until an initial processor qualified level is obtained based on the comparison.

[0076] An edge device location map is established based on the geographic locations of the qualified edge devices, and the initial processor qualification level of each qualified edge device is marked on the edge device location map accordingly.

[0077] The geographic location of a qualified edge device refers to its position in the three-dimensional space of the regional energy management scope. Based on the geographic location, a data visualization tool is used to map the processor qualification level of edge devices with qualified processor performance to the three-dimensional space of the regional energy management scope and form an edge device location map, wherein the edge device location map is classified according to the processor qualification level of the edge devices to form a location distribution.

[0078] Obtain a regional energy simulation model stored in a database, apply an initial multi-objective optimization scheme to the regional energy simulation model, obtain an energy efficiency evaluation value of the regional energy simulation model, and obtain a comprehensive energy efficiency difference value based on the energy efficiency evaluation value.

[0079] Obtain the energy efficiency evaluation value of the regional energy simulation model, and obtain the comprehensive energy efficiency difference value based on the energy efficiency evaluation value of the regional energy simulation model. The specific analysis process is: obtain the energy efficiency data set after the initial multi-objective optimization scheme is applied to the regional energy simulation model; obtain the correction factor data set set in the database; obtain the energy efficiency evaluation value based on the energy efficiency data set and the correction factor data set set in the database; obtain the energy efficiency target value stored in the database; calculate the comprehensive energy efficiency difference value based on the energy efficiency evaluation value, the energy efficiency target value and the energy efficiency difference evaluation model.

[0080] The calculation formula of the energy efficiency difference assessment model is:

[0081] ;

[0082] In the formula, is the comprehensive difference value of energy efficiency, is the energy performance evaluation value, is the energy efficiency target value stored in the database.

[0083] It is used to measure the degree to which the energy efficiency evaluation value deviates from the energy efficiency target value stored in the database. Its value not only takes into account and The numerical differences between and The greater the relative difference between them, the greater the impact on the energy efficiency evaluation value. It is the quantitative result of evaluating the overall regional energy efficiency quality based on the regional energy optimization result after the initial multi-objective optimization scheme is applied to the regional energy simulation model. The energy efficiency target value stored in the database corresponds to the initial multi-objective optimization solution, and is the standard energy efficiency that can be achieved under the ideal state of the initial multi-objective optimization solution.

[0084] Energy efficiency data set, specifically including power generation storage ratio, energy stability range, energy load rate, unit carbon emission intensity, and renewable energy ratio. Among them, the power storage ratio is the ratio of the amount of electricity that can be stored in the energy region to the amount of electricity generated. The energy stability range refers to the fluctuation range between energy supply and demand within an energy supply cycle. This indicator is expressed as the difference between the maximum power supply capacity and the minimum power supply capacity. The energy load rate refers to the ratio of actual power consumption within an energy supply cycle to the maximum rated power consumption within that time. The unit carbon emission intensity refers to the amount of carbon dioxide emissions generated when each unit of electricity (per kilowatt-hour) is generated. The renewable energy ratio refers to the proportion of renewable energy (such as solar energy, wind energy, etc.) in the total regional energy supply.

[0085] The correction factor data set includes the power generation and energy storage ratio correction factor, energy stability extreme correction factor, energy load rate correction factor, unit carbon emission intensity correction factor, and renewable energy ratio correction factor.

[0086] The specific process of obtaining the energy load rate is as follows: obtaining the energy load parameters after the initial multi-objective optimization scheme is applied to the regional energy simulation model, the energy load parameters including the peak-valley load difference, the power generation capacity factor, the grid frequency fluctuation rate, and the power supply redundancy level; storing the energy load parameters as an energy load designated tag, comparing the energy load designated tag with the energy load rates corresponding to the energy load designated tags stored in the database, and obtaining the energy load rate corresponding to the energy load designated tag.

[0087] The peak-valley load difference of electricity refers to the difference between the maximum load and the minimum load in an energy supply cycle. It reflects the volatility of electricity demand. A larger peak-valley difference means that the power grid needs to have a higher regulation capability. A large load difference may lead to an increase in the volatility of actual electricity consumption, which has a direct impact on the energy load rate. When the peak load is significantly higher than the average load, the load rate will decrease. The power generation capacity factor refers to the ratio between the actual power generation and the maximum possible power generation under the existing equipment capacity and operating time. It is used to measure the efficiency and utilization of power generation facilities. A higher power generation capacity factor means better regional energy utilization. If the power generation capacity factor is low, it means that the actual power generation is low and cannot meet the demand during peak periods, resulting in a decrease in the energy load rate.

[0088] The grid frequency fluctuation rate refers to the degree of fluctuation of the grid frequency within an energy supply cycle. The stability of the grid frequency is an important indicator of the power system. Frequency fluctuation reflects the stability of the balance between energy load supply and demand. High frequency fluctuation rate may lead to system instability, affect the actual power supply capacity, reduce the reliability of power supply, and further affect the energy load rate. The power supply redundancy level refers to the additional power generation capacity or backup capacity added to the power system to ensure reliable power supply. A good redundancy level can enhance the reliability of regional energy load operation during peak load periods. Since redundant power sources can continue to provide power when the main power generation facilities fail, the improvement of the power supply redundancy level can generally reduce the negative impact on the load rate, especially during peak demand periods.

[0089] The calculation formula for energy efficiency assessment value is:

[0090] ;

[0091] In the formula, k C k is the power generation energy storage ratio correction factor, J is the energy stability extreme correction factor, k F is the energy load rate correction factor, k Q is the unit carbon emission intensity correction factor, k Z is the correction factor for the proportion of renewable energy, C is the proportion of power generation and energy storage, J is the energy stability range, F is the energy load rate, Q is the unit carbon emission intensity, and Z is the proportion of renewable energy.

[0092] A higher C can store energy when electricity demand is low and release it during peak demand periods, which helps to increase F. Increasing C can balance the system load and reduce energy waste. At the same time, a higher C usually means that load fluctuations can be managed more effectively, thereby reducing J. Better C enhances the stability of the power grid. Since the power generation process of renewable energy produces almost no carbon dioxide, the higher the Z, the lower the Q. At the same time, a high Z usually directly leads to a decrease in Q. A higher F is usually associated with a smaller J because a high F means efficient energy utilization and can reduce the imbalance between supply and demand.

[0093] A higher C can effectively balance the production and consumption of electricity, avoid energy waste, and thus improve If C is high, it can effectively balance electricity demand and supply, increase F, and reduce J at the same time, indicating that regional energy can better cope with changes in power load, can support higher Z and increase F, and help stabilize energy operation, thereby improving , lower Q and increased Z means more reliance on low-carbon or zero-carbon energy sources, which means less environmental impact and cleaner energy use, and higher energy efficiency quality, Bigger.

[0094] If the energy efficiency comprehensive difference value is not greater than the energy efficiency comprehensive difference standard value stored in the database, the initial multi-objective optimization scheme is applied to the actual regional energy.

[0095] If the comprehensive energy efficiency difference value is greater than the standard value of the comprehensive energy efficiency difference stored in the database, the comprehensive energy efficiency difference value is used to match multiple pre-selected management optimization schemes corresponding to the initial multi-objective optimization scheme stored in the database, and they are applied one by one to the regional energy simulation model.

[0096] The comprehensive difference value of energy efficiency is compared with the standard value of comprehensive difference of energy efficiency stored in the database, and a dynamic adjustment mechanism based on the comprehensive difference value of energy efficiency is introduced to select different decision paths. This flexibility makes regional energy management more adaptable. By matching the initial multi-objective optimization scheme and the pre-selected management optimization scheme stored in the database, a data-driven decision-making process is realized, which improves the scientificity and accuracy of regional energy management decisions. At the same time, by combining the initial multi-objective optimization scheme with the regional energy simulation model, a closed-loop optimization verification process is formed, which can timely feedback the actual application effect and make adjustments, which helps to better match the actual regional energy demand with available resources and improve the overall regional energy optimization efficiency.

[0097] Obtain the comprehensive difference value of optimized energy efficiency after each pre-selected management optimization scheme is applied to the regional energy simulation model and the processor performance index of each edge device in the device site map.

[0098] The optimal management optimization scheme is determined from the pre-selected management optimization schemes based on the comprehensive difference value of energy efficiency after optimization and the processor performance index of each edge device in the device site map.

[0099] Determine the optimal management optimization plan, including the following steps: determine the edge device qualification rate based on the processor performance index of each edge device; obtain the processor qualification level of the qualified edge device; compare and analyze the processor qualification level of the qualified edge device with the initial processor qualification level to determine the qualification level change coefficient of the qualified edge device; comprehensively analyze the optimized energy efficiency comprehensive difference value, edge device qualification rate and qualified edge device qualification level change coefficient to obtain the comprehensive evaluation coefficient; determine the pre-selected management optimization plan corresponding to the smallest comprehensive evaluation coefficient as the optimal management optimization plan.

[0100] The edge device qualification rate is based on all edge devices within the scope of regional energy management, which is equal to the ratio of the number of qualified edge devices to the total number of edge devices. The qualified level change coefficient of qualified edge devices is based on qualified edge devices. The difference between the qualified level of the optimized qualified edge device processor and the qualified level of the initial processor is calculated, and the average is calculated for the difference. If the mean calculation result is a positive number, it means that the quality of the qualified edge device processor is further improved and the qualified quality is improved. On the contrary, if the mean calculation result is a negative number, it means that the overall performance level of the qualified edge device processor has declined. It is necessary to be vigilant whether it has become unqualified and its impact on the comprehensive evaluation coefficient. Integrated calculation at the edge to obtain the comprehensive evaluation coefficient can make quick decisions based on real-time data and determine the optimal management optimization plan.

[0101] The calculation formula of the comprehensive evaluation coefficient is:

[0102] ;

[0103] In the formula, is the comprehensive evaluation coefficient, CY is the comprehensive difference value of energy efficiency after optimization, is the edge device qualification rate, is the coefficient of variation of the qualified grade of the qualified edge devices.

[0104] It is used to indicate the comprehensive evaluation of the regional energy optimization effect for each pre-selected management optimization scheme. CY refers to the degree to which the energy efficiency evaluation value after applying the pre-selected management optimization scheme deviates from the energy efficiency target value stored in the database. By applying the pre-selected management optimization scheme, the reduction of CY indicates that the optimization scheme has successfully narrowed the gap between theoretical efficiency and actual efficiency, and improved the efficiency of regional energy management.

[0105] The implementation of the optimization scheme makes the system's energy consumption closer to the theoretical optimal level. Become smaller, Improvement means that the data processing performance and reliability of edge devices are enhanced, which can effectively support the efficient operation of regional energy management. Can reflect the improvement of edge device quality, When the comprehensive qualification level of edge devices is improved on the original qualification level, that is, it is always in a state of meeting the performance requirements of the device processor of regional energy management. right Further calculation and evaluation are performed when When the overall qualification level of edge devices decreases, the processor performance of edge devices decreases, and the overall risk of existing pre-selected management solutions cannot be borne. The impact of all edges on regional energy data processing should be considered to reduce the potential errors caused by the risks of existing pre-selected management solutions. Therefore, it should be based on right Further computational evaluation is performed.

Claims

1. Regional energy management optimization method based on edge computing and multi-objective optimization, characterized in that: The following steps are involved: The regional energy data collected within the regional energy management scope is sent from the data processing center to each edge device, the processor performance parameters of any edge device are obtained, and the edge devices with qualified processor performance are determined based on the processor performance parameters, recorded as qualified edge devices, and the edge devices with unqualified processor performance are replaced; Obtaining initial processor qualification levels for qualified edge devices; Establish an edge device location map based on the geographic locations of the qualified edge devices, and mark the initial processor qualification level of each qualified edge device on the edge device location map; Obtaining a regional energy simulation model stored in a database, applying an initial multi-objective optimization scheme to the regional energy simulation model, obtaining an energy efficiency evaluation value of the regional energy simulation model, and obtaining an energy efficiency comprehensive difference value based on the energy efficiency evaluation value; If the energy efficiency comprehensive difference value is not greater than the energy efficiency comprehensive difference standard value stored in the database, the initial multi-objective optimization scheme is applied to the actual regional energy; If the energy efficiency comprehensive difference value is greater than the energy efficiency comprehensive difference standard value stored in the database, multiple pre-selected management optimization schemes corresponding to the initial multi-objective optimization scheme stored in the database are matched by the energy efficiency comprehensive difference value, and applied to the regional energy simulation model one by one; Obtain the optimized comprehensive difference value of energy efficiency after each pre-selected management optimization scheme is applied to the regional energy simulation model and the processor performance index of each edge device in the device site map; Determine the best management optimization scheme from the pre-selected management optimization schemes based on the comprehensive difference value of energy efficiency after optimization and the processor performance index of each edge device in the device site map; The specific analysis process of determining edge devices with qualified processor performance based on processor performance parameters is as follows: Obtain the processor performance parameters of any edge device within the regional energy management range, including clock frequency deviation value, thread average response rate, and instruction hit rate; Based on the processor performance parameters, the compensation factor parameters set by the edge device in the database are obtained, where the compensation factor parameters include a clock frequency deviation value compensation factor, a thread average response rate compensation factor, and an instruction hit rate compensation factor; Obtaining a processor performance index of the edge device based on the processor performance parameter and the compensation factor parameter; Extract the edge device processor performance standard value stored in the database, and compare the processor performance index with the edge device processor performance standard value: If the processor performance index of the edge device is less than the edge device processor performance standard value, the edge device processor is marked as unqualified; If the processor performance index of the edge device is not less than the standard value of the edge device processor performance, the edge device processor is marked as qualified and determined as an edge device with qualified processor performance; The calculation formula of the processor performance index of the edge device is as follows: ； In the formula, α is the performance index of the edge device processor, p is the clock frequency deviation value, s is the average thread response rate, m is the instruction hit rate, and k is the average thread response rate. p is the clock frequency deviation compensation factor set in the database, k s k is the average response rate compensation factor of threads set in the database, m is the instruction hit rate compensation factor set in the database, and e is a natural constant; The energy efficiency evaluation value of the regional energy simulation model is obtained, and the energy efficiency comprehensive difference value is obtained based on the energy efficiency evaluation value of the regional energy simulation model. The specific analysis process is as follows: Obtaining an energy efficiency data set after the initial multi-objective optimization scheme is applied to the regional energy simulation model; Obtaining a correction factor data set set in a database; Obtaining an energy efficiency evaluation value based on an energy efficiency data set and a correction factor data set set in a database; Obtaining energy efficiency target values ​​stored in a database; Calculate the comprehensive difference value of energy efficiency based on the energy efficiency evaluation value, the energy efficiency target value and the energy efficiency difference evaluation model; The energy efficiency data set specifically includes the power generation and energy storage ratio, energy stability extreme difference, energy load rate, unit carbon emission intensity, and renewable energy ratio; The correction factor data set includes the power generation energy storage ratio correction factor, energy stability extreme correction factor, energy load rate correction factor, unit carbon emission intensity correction factor, and renewable energy ratio correction factor; The specific process of obtaining the energy load rate is as follows: Obtain energy load parameters after the initial multi-objective optimization scheme is applied to the regional energy simulation model, the energy load parameters including power peak-valley load difference, power generation capacity factor, grid frequency fluctuation rate, and power supply redundancy level; The energy load parameter is stored as an energy load designated tag, and the energy load designated tag is compared with the energy load rate corresponding to each energy load designated tag stored in the database to obtain the energy load rate corresponding to the energy load designated tag; The calculation formula of the energy efficiency evaluation value is: ； In the formula, is the energy efficiency evaluation value, k C k is the power generation energy storage ratio correction factor, J is the energy stability extreme correction factor, k F is the energy load rate correction factor, k Q is the unit carbon emission intensity correction factor, k Z is the correction factor for the proportion of renewable energy, C is the proportion of power generation and energy storage, J is the energy stability range, F is the energy load rate, Q is the unit carbon emission intensity, Z is the proportion of renewable energy, and e is a natural constant.

2. The regional energy management optimization method based on edge computing and multi-objective optimization according to claim 1 is characterized in that: The specific analysis process of obtaining the initial processor qualification level of the qualified edge device is as follows: Calculate the processor performance deviation rate based on the processor performance index of the qualified edge device and the edge device processor performance standard value stored in the database; Extracting the processor performance deviation rate grading interval endpoints stored in the database, the processor performance deviation rate grading interval endpoints stored in the database are the basis for grading the processor qualification level; The processor qualification level is graded based on the processor performance deviation rate and the processor performance deviation rate grading interval endpoints stored in the database to obtain an initial processor qualification level.

3. The regional energy management optimization method based on edge computing and multi-objective optimization according to claim 1 is characterized in that: The method of determining the optimal management optimization scheme from the pre-selected management optimization schemes based on the optimized energy efficiency comprehensive difference value and the processor performance index of each edge device in the device site map comprises the following steps: Determine the edge device qualification rate based on the processor performance index of each edge device; Obtain the processor qualification level of the qualified edge device; Compare and analyze the processor qualification level of the qualified edge device with the initial processor qualification level to determine the qualification level change coefficient of the qualified edge device; Comprehensively analyze the comprehensive difference value of energy efficiency after optimization, the qualified rate of edge devices, and the qualified grade variation coefficient of qualified edge devices to obtain a comprehensive evaluation coefficient; The pre-selected management optimization plan corresponding to the smallest comprehensive evaluation coefficient is determined as the optimal management optimization plan.

4. The regional energy management optimization method based on edge computing and multi-objective optimization according to claim 3 is characterized in that: The calculation formula of the comprehensive evaluation coefficient is: ； In the formula, is the comprehensive evaluation coefficient, CY is the comprehensive difference value of energy efficiency after optimization, is the edge device qualification rate, is the coefficient of variation of the qualified grade of qualified edge devices, and e is a natural constant.

Images