CN-122000914-A - Power distribution area shared energy storage optimization scheduling method with cooperation of multiple elastic resources

Abstract

The invention discloses a method for optimally scheduling shared energy storage of a distribution area with cooperation of multiple elastic resources, which comprises the steps of bringing four core resources of distributed photovoltaics, electric vehicle charging station loads, temperature control loads and shared energy storage batteries into a unified scheduling system, establishing a system decision model, setting a multidimensional constraint condition, forming a scheduling closed loop with cooperation of source, storage and load, effectively stabilizing photovoltaic output fluctuation, introducing 'unit carbon emission cost', indirectly converting carbon emission into carbon cost into a comprehensive cost objective function, bringing fossil energy supplementing carbon emission and energy storage charge and discharge indirect carbon emission caused by insufficient photovoltaic in the area into cost items, constructing a layered iterative optimization mechanism, innovatively adopting a layered optimization framework of 'user side-platform side', dynamically adjusting energy storage price and regulating air conditioner loads by combining a user scheme, generating a Pareto optimal curve, and realizing benefit balance between users and a power grid. The method comprises the steps of constructing a full-flow mechanism of 'real-time data acquisition-collaborative computing platform analysis-abnormal reporting-regulation and control instruction issuing', effectively avoiding the problem that the traditional 'static scheduling' cannot cope with sudden working conditions in real-time response, ensuring that a photovoltaic output fluctuation and load mutation scene system can still stably operate, and improving the power supply reliability and resource utilization efficiency of a platform region.

Inventors

• ZHANG MIN
• XU XINYU
• LI JUN
• DAI LIYUAN

Assignees

• 国网江苏省电力有限公司南通供电分公司

Dates

Publication Date

20260508

Application Date

20251201

Claims (6)

1. 1. The energy storage optimization scheduling method for the power distribution station sharing with the cooperation of multiple elastic resources is characterized by comprising the following steps of: s1, constructing a photovoltaic curve and initializing basic parameters; S1-1, photovoltaic data acquisition and curve generation, namely acquiring historical operation data and real-time output power of all distributed photovoltaic in a platform region, determining a day-night photovoltaic output curve (distinguishing a peak output period in daytime and a zero output period in nighttime) of the current platform region through data fitting, simultaneously adopting a photovoltaic prediction algorithm (such as a machine learning prediction model based on meteorological data), calculating an increment curve of photovoltaic output in the future 24 hours, and determining a fluctuation range and a peak node of photovoltaic output in each period; S1-2, setting a time dimension and basic parameters, namely dividing a 24h control period into 96 regulation and control periods, wherein the time interval delta t=15 min of each period, namely, the control period T= {1,2, ⋯,96}, and simultaneously initializing key basic parameters, wherein the key basic parameters comprise charge and discharge reference prices of shared energy storage, electricity price intervals when a station is distinguished, and unit carbon emission cost such as the determination based on regional carbon tax standards or carbon quota transaction prices; S2, constructing a system decision model, namely constructing a system decision model covering 'load-energy storage-photovoltaic' cooperation based on the characteristics of multiple resources in a platform region, and determining the optimization targets and the operation characteristics of the resources of each main body: S2-1, quantifying resource characteristics: The electric vehicle charging station load comprises the steps of counting the time interval distribution of the charging demands of users, such as commuting, taking 18:00-22:00 as charging peaks, determining the upper limit of single-pile charging power, the charging duration demands and the load duty ratio capable of being scheduled in a staggered mode; The shared energy storage battery is used for defining the charge and discharge efficiency of energy storage, To charge efficiency, For discharging efficiency, rated capacity S2-2, embedding a two-dimensional objective function, wherein the SOC safety interval of the charge state is more than or equal to 0 and less than or equal to 1, and the maximum charge and discharge power limit corresponding to different SOC intervals is shown as follows: The user side objective function takes 'the minimum user comprehensive cost' as a core, and the cost items comprise the running cost of electric equipment and the time-of-use electricity price of electricity purchasing cost Calculating and sharing the energy storage use cost, innovatively adding a carbon emission cost item, and quantitatively calculating indirect carbon emission based on the electric quantity of non-clean energy used by a user and energy storage charge and discharge; The objective function at the side of the platform area takes 'peak load reduction + distributed photovoltaic complete absorption' as a core, and simultaneously associates peak clipping cost, comprises user comfort level loss compensation and energy storage Gu Chongfeng gain loss, and forms an optimization direction of 'technical index-economic cost-low carbon target' cooperation with carbon emission control target limiting carbon emission of fossil energy supply when peak time; s3, setting multi-dimensional constraint conditions, and guaranteeing scheduling safety and low carbon property; four types of core constraint conditions are set by combining the running requirement of a power grid of a platform region, the physical characteristics of resources and the carbon emission control requirement: s3-1, photovoltaic power generation power constraint, namely real-time photovoltaic output of user i Needs to meet the requirements of The upper limit of the photovoltaic output of the user i is set, meanwhile, the total photovoltaic output of the platform area is matched with the load and the energy storage charge-discharge power, so that carbon emission loss caused by light abandoning is avoided, namely the photovoltaic output is preferably used or stored, and direct waste is avoided; S3-2, elastic load and energy storage constraint: The temperature control load is that the running power of the air conditioner is required to be within the rated power range of the equipment, and the temperature adjustment frequency is not more than a user comfort threshold value, for example, the starting and stopping time is not more than 1 time per hour; The electric vehicle is charged, wherein the charging power of a single pile does not exceed the rated power of a charging pile, and the basic charging requirement of a user is ensured when peak shifting scheduling is carried out, for example, the charging time is not less than the minimum requirement time; Sharing energy storage: the charging and discharging power of the energy storage of the user side needs to be satisfied The negative sign is discharge, and the total energy storage charge and discharge power of the station area needs to be satisfied Avoiding overload of energy storage and maintaining the SOC of energy storage And at the end of the scheduling period The initial safety value is required to return, so that the next-day energy storage and power shortage are avoided; s3-3, the electricity purchasing constraint of the user is that the electricity purchasing power of the user i Needs to meet the requirements of The negative sign is electricity selling, and the service charge is required to be deducted when electricity is sold Avoiding excessive electricity selling of users influence the stability of the power grid; s3-4, limiting the total carbon emission amount through indirect parameters, wherein the energy supplementing electric quantity of fossil energy is not more than 10% of the total load in a peak of a platform area, and the carbon emission intensity in the energy storage charging and discharging process is not more than the regional low-carbon standard; S4, real-time data acquisition and collaborative computing platform monitoring analysis; s4-1, full-dimensional data acquisition and transmission, namely acquiring the following data in real time through equipment such as a smart electric meter, a photovoltaic inverter, an energy storage monitoring terminal, a load controller and the like in a platform region and transmitting the data to a cooperative computing platform: Photovoltaic power generation data, namely real-time output, accumulated power generation amount and predicted deviation value of each photovoltaic module; load operation data, namely real-time power and temperature set values of temperature control loads, charging pile operation states and power of electric vehicle charging stations and period data of non-adjustable resident base loads; The energy storage equipment data is shared with the real-time SOC, the charge and discharge power, the charge and discharge efficiency and the equipment fault information of energy storage; Cost and carbon emission data, namely real-time-sharing electricity price, unit carbon emission cost, electricity purchasing cost and carbon emission cost generated by a user; s4-2, platform data analysis and anomaly identification, wherein the collaborative computing platform performs three core processes on the acquired data: The compliance verification comprises the steps of comparing data with S3 constraint conditions, and identifying abnormal data deviating from the constraint, such as that the energy storage SOC is lower than 0.2, the photovoltaic output exceeds the upper limit, and the charging power of the electric vehicle is overloaded; Calculating total carbon emission of a station area in the current period, and marking the total carbon emission as abnormal data if the total carbon emission exceeds a low-carbon target threshold value based on energy supplementing electric quantity of fossil energy and carbon emission of stored energy charge and discharge; The optimality evaluation, namely analyzing the power matching degree of the current load-photovoltaic-energy storage, identifying the problems of photovoltaic light rejection, excessive load at peak time and idle energy storage equivalent rate, and generating a node report to be optimized; s4-3, reporting abnormal data, namely reporting data classification of compliance abnormality, carbon emission abnormality and efficiency abnormality to a platform area management end, and simultaneously automatically triggering an early warning mechanism to ensure timely response of abnormal problems; s5, issuing a regulation instruction and executing the regulation instruction in cooperation with multiple resources; Based on the analysis result of the collaborative computing platform, combining with layered logic of 'user side iterative optimization-platform side peak clipping optimization', issuing regulation and control instructions step by step and executing: S5-1, executing a user side optimization instruction, namely issuing an optimization instruction based on 'the minimum comprehensive cost and the carbon-containing cost' to each user terminal, and guiding a user to adjust electricity utilization and energy storage strategies: the photovoltaic output peak time is 10:00-15:00, for example, the user is instructed to use the photovoltaic spontaneous self-use preferentially, the power grid electricity purchasing is reduced, and meanwhile, the shared energy storage is controlled to enter a charging mode, and the low-price electricity and low-carbon energy source are utilized for charging; The electricity consumption peak and the carbon emission peak time are respectively 18:00-21:00, after the peak shifting charge of the electric vehicle charging station is instructed to be transferred to 22:00, the temperature setting value of the air conditioner is adjusted to be 1-2 ℃ in summer, and the energy storage is shared to enter a discharging mode to supplement a load gap, so that fossil energy supplementing is reduced; S5-2, peak clipping and low-carbon regulation at the side of the platform region, wherein the platform region management terminal further performs cooperative regulation and control aiming at the peak load time based on the execution result at the user side: The temperature control load cluster regulation and control comprises the steps of carrying out batch scheduling on air conditioning loads which are intensively operated in a platform region, reducing peak time loads such as starting and stopping of air conditioning of each building in a time-division manner in a 'wheel stopping' or 'power reduction' mode, and simultaneously ensuring that the comfort loss of users is minimum; The shared energy storage cooperative peak regulation comprises the steps of regulating a charging and discharging plan of public energy storage at the side of a platform region, if the load exceeds the upper limit during peak time, commanding the public energy storage to increase the discharging power, and if the photovoltaic output is excessive, commanding the public energy storage to prolong the charging time length and avoid discarding light; S5-3, performing closed loop verification on the regulation effect, namely feeding back the regulated power generation power, the photovoltaic energy supplementing power supply, the stored power shared energy storage and the load power data to a collaborative computing platform in real time after the instruction is executed by each platform area, and verifying whether three targets of 'complete photovoltaic absorption, compliance of constraint conditions and minimum carbon emission cost' are met by the platform; Converting carbon emission into carbon cost, embedding the carbon cost into a target function of a user side-station area side, realizing resource data support by depending on VPP, The user side objective function construction is that the user is definitely aimed at the 'minimum comprehensive cost', and the mathematical expression is as follows: (1) in the formula (1), the components are as follows, For the integrated cost of the user i, In order to purchase the cost of electricity, In order to share the cost of energy storage usage, As a matter of carbon-emission costs, Is the cost per unit of carbon emission, Carbon emissions for the user t period; the objective function of the platform side is related to the platform side taking 'peak load reduction plus carbon emission control' as a target, and the mathematical model is as follows: (2) (3) In the formulas (2) and (3), J is a comprehensive optimization objective function value; is the total load of the platform area; the renewable energy source output is used for the station area; in order to adjust the energy storage charging and discharging power of the foreground area side, the value is positive to represent charging and negative to represent discharging; Peak clipping cost for the platform area side; Is a weight coefficient; the amount of power change for user i; The charging and discharging power of the energy storage equipment at the side of the platform area is changed due to peak clipping requirements; Peak clipping costs at the site side include loss of revenue from energy storage Gu Chongfeng for adjusting elastic load at the user side, which can be expressed specifically as: (4) in the formula (4): For the retail electricity prices of the areas, The state of charge of the energy storage at the cell side can be expressed as: (5) Wherein: the charge state of energy storage at the zone side of the time period t; Charging and discharging power for energy storage at the side of the zone of the period t; charging and discharging efficiencies of energy storage at the side of the platform area respectively; the energy storage capacity of the platform area side; distributed photovoltaic in polymerization platform region depending on VPP platform real-time operational data of elastic load and energy storage device (e.g 、 ) As in the objective function The parameters such as the parameters provide accurate input, so that the accuracy of cost quantization and optimization calculation is ensured; the carbon emission control is used as a guide, and a CPLEX optimization algorithm and VPP data interaction layering iteration optimization mechanism are combined, wherein the specific flow and mathematical constraint are as follows: and (3) user-side iterative optimization, namely solving a user-side objective function by adopting a CPLEX tool, and simultaneously meeting multiple constraint conditions, wherein the key mathematical constraint comprises: Power balance constraint: ; Energy storage SOC constraints: ; carbon emission implicit constraint: , the upper limit of the carbon emission coefficient of unit electricity purchase; solving the objective function with constraint through CPLEX to generate a strategy of optimal electricity consumption and energy storage charge and discharge of a user, such as preferentially absorbing photovoltaic to reduce , The optimization of the peak clipping of the side of the platform area, namely, solving a target function of the side of the platform area based on an optimization result of the side of the user and generating a Pareto optimal front curve of peak clipping cost, peak load and carbon emission, wherein the key is that the power deviation of the temperature control load is regulated Deviation from energy storage power of a cell The energy supplementing carbon emission of fossil energy during peak time is ensured not to exceed 10% of the total load; VPP assisted completion of user-side optimization results, e.g Interaction transmission with a platform-side regulation instruction is carried out, iteration optimization deviation caused by data island is avoided, and the combination of a layered iteration mechanism and a CPLEX solving algorithm is a core operation path for landing a carbon emission target; The data integration is realized by means of VPP, and a data acquisition-constraint verification technology system for supporting carbon emission quantification is provided, which comprises the following steps: And the multidimensional data acquisition specification is that the real-time data of multiple devices in the platform area are integrated through the VPP platform, and the core data to be acquired and the mathematics are related as follows: photovoltaic data: is a real-time photovoltaic output, For upper limit of photovoltaic output, for accounting Carbon emission reduction of the photovoltaic alternative electricity purchasing, And (3) energy storage data: Is charged and discharged with power, Is in a charged state, For charge and discharge efficiency, the method is used for calculating the indirect carbon emission of energy storage charge and discharge: , for the energy storage carbon emission coefficient, Load data: Is the temperature control/electric vehicle load power, For load regulation deviation, is used for checking the satisfaction condition of the implicit constraint of carbon emission, Constraint checking mathematical logic, wherein the VPP platform automatically checks three core constraints based on the acquired data: Photovoltaic output constraint: The carbon emission loss caused by light discarding is avoided; energy storage charge-discharge power constraint: low-efficiency carbon emission caused by energy storage overload is avoided; total carbon emission constraint: the upper limit of carbon emission of the stage section is set; the unified collection and constraint verification of the data are realized through the VPP, and a basic guarantee is provided for quantifying the carbon emission cost and optimizing the target landing.
2. 2. The method is characterized in that a system decision model integrating multiple elastic resource characteristics is constructed, and a unified system decision model is constructed for integrating and modeling three types of elastic resource characteristics of temperature control loads such as air conditioners, electric vehicle charging station loads and shared energy storage batteries in a residential area, so that the method is a basic support for carbon emission consideration optimal scheduling, and in step S2, the cooperative integration of multiple resource characteristics is realized by quantitatively analyzing the core operation characteristics of the three types of resources, namely the correlation characteristics of the temperature control loads, such as the corresponding relation between the start and stop threshold of the air conditioners and the operation power, the characteristic of the charge demand period of the electric vehicle charging stations, such as the power concentration requirement of the peak charge period, and the characteristic of the charge and discharge efficiency of the shared energy storage batteries, such as the upper limit difference of charge and discharge power of different SOC intervals, and converting the operation boundary and constraint conditions of the three types of resources into a computable parameterized model.
3. 3. The method for optimizing and dispatching the shared energy storage of the distribution area with the cooperation of multiple elastic resources according to claim 2 is characterized in that a layered solution flow of 'iterative optimization at a user side-peak clipping optimization at the area side' aiming at realizing the cooperation of 'economic cost-power grid efficiency-carbon emission reduction' is a key operation path of a core carbon emission optimization target landing, the specific flow corresponds to three core steps of a solution stage, namely, the first stage finishes user load, distributed photovoltaic output prediction and shared energy storage reference price initialization, provides a data basis for subsequent optimization, the second stage aims at 'comprehensive cost of carbon cost is minimum', a CPLEX optimization tool is used for generating a power consumption and shared energy storage charging and discharging scheme, a platform management end dynamically adjusts the shared energy storage charging and discharging price according to the scheme, such as guiding a user to store energy charge in a low carbon period, and the third stage aims at 'distributed photovoltaic complete absorption and peak load reduction', controls a temperature control load running state in combination with the user scheme, and generates a 'peak clipping cost-peak load' Pareto optimal front curve, and finally determines a multi-target execution scheme. The layering iteration process ensures the cooperative realization of the core carbon emission target and the economic and power grid target through the bidirectional interaction of 'autonomous optimization of users-cooperative regulation and control of the transformer areas', and has uniqueness and operability.
4. 4. The method is characterized by comprising a real-time data-driven abnormal monitoring and dynamic regulation mechanism, and is characterized in that step S4 is specifically implemented by collecting photovoltaic power generation power and energy storage equipment operation data in the distribution area in real time, namely charging and discharging power, SOC (system on chip), efficiency and user load data, namely temperature control load and electric vehicle charging load, transmitting the data to a cooperative computing platform, automatically identifying abnormal data deviating from constraint, such as carbon emission risk caused by energy storage overcharge and photovoltaic dip by comparing preset constraint conditions, such as upper limit of photovoltaic power, limit of energy storage charging and discharging power and SOC safety interval, and reporting the abnormal data in real time, and based on an abnormal analysis result, sending regulation instructions to each area by the platform, adjusting power generation, such as limiting non-clean energy supplement and storage power, such as adjusting an energy storage strategy, ensuring the operation within a regression constraint range of the system, avoiding the defect that static scheduling cannot cope with a sudden working condition, and providing operation guarantee for stable execution of a core carbon emission optimization strategy.
5. 5. The optimal scheduling method for sharing energy storage in a power distribution area with cooperation of multiple elastic resources according to claim 4, wherein the multi-dimensional constraint design of the charge and discharge of the shared energy storage is used for supporting a multi-dimensional constraint condition design scheme of carbon-containing emission consideration scheduling aiming at the shared energy storage equipment, and the dual constraint of a user side and an area side is covered. In the constraint of the user side, the maximum charge and discharge power constraint of the stored energy is definitely shared ) Constraint of overall charge and discharge power of station area ) SOC constraint ) Constraint of charge and discharge efficiency ) In the side constraint of the platform region, the SOC constraint of the beginning and ending time period of energy storage is further supplemented, such as the SOC is maintained in a reasonable interval at the end of a dispatching period, the charging and discharging constraint of the next-day carbon emission increase and the linkage with photovoltaic absorption is avoided, such as the preferential energy storage and charging at the photovoltaic output peak, and the carbon emission loss caused by light abandoning is reduced.
6. 6. The energy storage optimization scheduling method for the power distribution station area sharing with the cooperation of multiple elastic resources according to claim 5, wherein historical operation data and real-time output power of distributed photovoltaic in the station area are obtained, and a current day and night photovoltaic output curve is determined through data fitting; calculating a future photovoltaic output increment curve by adopting a prediction algorithm, and determining a photovoltaic output fluctuation range and a peak node; dividing a control period into a plurality of regulation and control periods, and initializing a shared energy storage charging and discharging reference price, a power price interval and a unit carbon emission cost during a station division; Constructing a system decision model, and quantifying temperature control load temperature adjustment threshold correlation characteristics, electric vehicle charging station load period distribution characteristics and a shared energy storage battery state of charge safety interval; Embedding a user side objective function and a platform side objective function, wherein the user side objective function comprises electricity purchasing cost, shared energy storage using cost and carbon emission cost, and the platform side objective function comprises peak load reduction and carbon emission control; Setting photovoltaic power generation power constraint, elastic load and energy storage constraint, user electricity purchase constraint and carbon emission recessive constraint, wherein the photovoltaic power generation power constraint requires that the total photovoltaic output of a platform region is matched with the load and the energy storage charge and discharge power, and the elastic load and the energy storage constraint require that an energy storage charge state is shared to maintain a safe interval; Collecting photovoltaic power generation data, load operation data, energy storage equipment data and cost carbon emission data in real time through equipment, and transmitting the data to a cooperative computing platform, wherein the cooperative computing platform performs compliance verification, low-carbon analysis and optimality evaluation on the collected data, identifies abnormal data and forms an abnormal identification result; Generating a user side optimization instruction and a platform side regulation instruction based on the anomaly identification result and the system decision model, wherein the user side optimization instruction guides spontaneous self-use and shared energy storage charging in a light Fu Gaofeng period, and the platform side regulation instruction adjusts a temperature control load round stop and shared energy storage charging and discharging plan according to a load peak period; issuing the user side optimization instruction and the platform side regulation instruction to corresponding terminals for execution, acquiring the generated power, the stored power and the load power data after the execution, and feeding back to the collaborative computing platform; Verifying whether the data after execution meet the photovoltaic complete digestion and constraint condition compliance or not through the cooperative computing platform, judging a verification result, and if not, iteratively adjusting an optimization instruction to form a final scheduling scheme; And updating the charge-discharge price and load regulation strategy of the shared energy storage according to the final scheduling scheme, generating an optimized front curve corresponding to peak clipping cost and peak load, and determining a comprehensive optimization target value of the platform region.

Description

Power distribution area shared energy storage optimization scheduling method with cooperation of multiple elastic resources Technical Field The invention relates to the technical field of information, in particular to a power distribution station sharing energy storage optimization scheduling method with cooperation of multiple elastic resources. Background In the scene of the cooperative optimization of the photovoltaic and the shared energy storage in the area, the photovoltaic has the characteristics of small single machine, large quantity, wide distribution, irregularity, and the like, so that the problems of serious power flow foldback, node voltage out-of-limit and the like are easily caused after the photovoltaic is connected into a power distribution network, and meanwhile, the power network is difficult to realize the real-time supply and demand balance due to the strong fluctuation of the photovoltaic output, so that higher requirements are put forward on the flexibility and the resource cooperativity of a regulation strategy. How to accurately construct and dynamically adjust objective functions at the user side and the platform side under multi-objective constraints to balance optimal solutions for photovoltaic digestion, carbon emission control, and economic cost. The distributed photovoltaic output in the platform area has obvious daytime volatility, the night output is almost zero, the load demand is often misplaced between the peak period and the photovoltaic output peak period, so that the photovoltaic electric quantity cannot be completely consumed, and meanwhile, the charge and discharge strategy of the shared energy storage is required to give consideration to the electricity purchasing cost of a user and the reduction demand of the peak load of the platform area on the basis of maintaining the safe interval of the state of charge of the battery. In addition, the temperature regulation threshold value of the temperature control load and the time interval distribution characteristic of the electric vehicle charging station are highly dynamic, so that the complexity of regulation and control time interval division and instruction generation is increased. The implicit constraint requirement of carbon emission limits the energy supplementing proportion of fossil energy in the peak period, but when the existing data acquisition and cooperative calculation platform processes real-time data, abnormal data are difficult to quickly identify, low carbon property and optimality are accurately estimated, so that generated optimization instructions possibly deviate from actual requirements, and the photovoltaic absorption rate and the system stability are affected. More importantly, the initialization parameters of the electricity price interval when the energy storage charging and discharging reference price and the station are shared lack of an adaptive adjustment mechanism, and the prediction deviation of the photovoltaic output increment curve and the dynamic balance requirement under a load abrupt change scene are difficult to deal with. The current distribution area optimizing and dispatching strategy focuses on peak-staggering regulation and control of adjustable loads such as electric vehicle loads and temperature control loads (such as air conditioners), the core targets are concentrated on reducing economic cost of users, stabilizing peak-valley difference of loads or improving running stability of a power grid, but two key technical limitations generally exist, namely on one hand, on the other hand, the existing strategy does not fully consider energy storage battery resources connected to a power distribution network, photovoltaic volatility can not be relieved jointly from a source side and a storage side through multi-resource cooperation of a distributed photovoltaic-adjustable load-energy storage battery, so that dispatching effect is limited, on the other hand, on the premise of ' carbon reaching peak, carbon neutralization and ' policy background ', the existing strategy does not incorporate carbon emission into an optimizing system, namely, on the other hand, the existing strategy does not use the cooperation of ' carbon emission cost ' (such as cost accounting based on carbon tax and carbon quota transaction), electricity purchasing cost, energy storage use cost and the like as optimization targets, and on the other occasions of a scene design adaptation scheme of carbon emission indexes needing to be controlled for a low-carbon demonstration park, a zero-carbon community and the like, so that dispatching result is difficult to meet regional carbon emission control requirements and has a disjointing with the current energy development direction. The problem relates to the deep coupling of multidimensional data fusion, real-time anomaly identification and multi-objective iterative optimization in technical realization, and the limitation of the existing model and algorithm is urgen