Energy asset portfolio optimization analysis method, system and equipment based on three-dimensional risk conduction and medium

Abstract

The invention provides an energy asset investment portfolio optimization analysis method, system and device based on three-dimensional risk conduction and a medium, and belongs to the technical field of energy finance and intelligent decision making. The method comprises the steps that multi-source heterogeneous data is collected and comprises electric power data, meteorological data, industry standard text data and electric power equipment state data; performing data cleaning, space-time alignment processing and multi-source heterogeneous data fusion processing on the multi-source heterogeneous data; constructing a three-dimensional risk conduction model for performing industry standard sensitivity, meteorological risk quantification and power market fluctuation analysis; and according to the output result of the three-dimensional risk conduction model, constructing a multi-objective optimization framework, and by taking the maximization of the difference value between the expected income and the risk adjustment item as the objective, generating an optimal weight scheme of the energy asset portfolio in combination with the constraint condition. According to the invention, low-threshold and high-timeliness investment decision and risk management are realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of energy finance and intelligent decision-making, and particularly relates to an energy asset investment portfolio optimization analysis method, system, device and medium based on three-dimensional risk transmission. BACKGROUND

[0002] The current global energy transformation process is accelerating, and the proportion of new energy assets such as wind power and photovoltaic in the investment portfolio is continuously increasing. Energy asset investment decisions have shifted from single income orientation to multi-dimensional balance. However, existing energy asset investment portfolio optimization methods are difficult to adapt to the complex characteristics of new energy assets, and have significant technical limitations: Existing solutions focus on single-dimensional risks of power market price fluctuations or only consider the impact of meteorological factors in isolation, without constructing three-dimensional risks of industry standard dynamic changes, meteorological disasters and market fluctuations. Risk quantification relies on static models, which cannot calculate dynamic risk factors in real time, cannot reflect the coupling and superposition effects of multiple risk factors, and result in large risk assessment deviations. Traditional multi-objective optimization frameworks often ignore carbon emission limits and power grid transmission capacity constraints under the double carbon target, and optimization algorithms mostly use serial computing mode without combining distributed reinforcement learning and other technologies to improve solution efficiency. It is difficult to quickly generate a globally optimal asset weight scheme under complex constraints, ultimately resulting in insufficient scientificity and operability of investment decisions, and failing to meet the fine management needs of new energy asset investment portfolios. SUMMARY

[0003] To solve the problems of data island aggravating risk assessment distortion, large risk assessment deviation of static models, and calculation bottleneck of traditional algorithms limiting dynamic adjustment capability in the prior art, the application provides an energy asset investment portfolio optimization analysis method, system, device and medium based on three-dimensional risk transmission.

[0004] In a first aspect, the embodiments of the application provide an energy asset investment portfolio optimization analysis method based on three-dimensional risk transmission, comprising the following steps: S1. Collecting multi-source heterogeneous data including power data, meteorological data, industry standard text data and power equipment state data, performing data cleaning, spatio-temporal alignment processing on the multi-source heterogeneous data, and performing fusion processing on the processed multi-source heterogeneous data; S2. Constructing a three-dimensional risk transmission model for industry standard sensitivity, meteorological risk quantification and power market fluctuation analysis; S3. According to the output result of the three-dimensional risk transmission model, constructing a multi-objective optimization framework, taking the difference between the expected return and the risk adjustment term as the target, and generating an optimal weight scheme of the energy asset investment portfolio under the constraint condition.

[0005] Further, in step S1: The power data includes node price and transaction information of a power market, the meteorological data includes historical disaster satellite image data, and the power equipment state data includes SCADA (Supervisory Control And Data Acquisition) data and life model data. The data cleaning uses an isolation forest algorithm to filter abnormal data, and the spatio-temporal alignment uses an Apache Kafka streaming processing platform for timestamp alignment.

[0006] Further, in step S1, the fusion processing of the processed multi-source heterogeneous data includes: integrating the multi-source data after data cleaning and spatio-temporal alignment into standardized spatio-temporal data through a unified data format and structure.

[0007] Further, the specific steps of step S2 are: S21. Using a BERT model to perform semantic analysis on industry standard text data, extracting key constraint factors including technical specifications, environmental protection requirements, and safety standards, establishing a mapping relationship between the constraint factors and asset return rate, constructing an influence function of industry standard factors, and outputting a correction coefficient of asset return rate; S22. Using historical disaster satellite image data to train a U-Net convolutional neural network, constructing a prediction model of wind power icing damage and photovoltaic dust shielding, obtaining a 1kM resolution wind power icing level probability map and a photovoltaic dust shielding probability map of the power equipment through the prediction model, and mapping the wind power icing level probability map and the photovoltaic dust shielding probability map into an influence function of meteorological risk factors, and outputting an availability attenuation coefficient; S23. Using an improved GARCH-M model, incorporating conditional volatility into the mean equation, inputting power data and power equipment data into the improved GARCH-M model, outputting conditional mean and conditional volatility, constructing an influence function of power market volatility factors, and outputting an asset risk correction coefficient; S24. Coupling the influence functions of industry standard factors, meteorological risk factors, and power market volatility factors to calculate dynamic risk factors, using a conditional risk value CVaR+ model to calculate the risk value CVaR under a confidence level α every 5 minutes α :

[0008] where R i is a dynamic risk factor, γ is a risk value under a confidence level α, and p i is the occurrence probability of the i-th asset.

[0009] Further, in step S24, the calculation formula of the dynamic risk factor R i is:

[0010] Among them, R 0,i For the benchmark return, β i θ i φ i These represent the impact weights of industry standards, weather risks, and electricity market volatility, used to measure the degree of impact of different factors on asset returns; S i M is the output of the influence function of industry standard factors. i E is the output of the influence function of meteorological risk factors. i This is the output of the function representing the impact of electricity market fluctuations.

[0011] Further, in step S3, the objective function of the multi-objective optimization framework is:

[0012] Among them, w i Let E(R) be the weight of the i-th asset class. i ) represents the expected return, λ represents the risk aversion coefficient, and CVaR α Value at risk at confidence level α; The constraints include: a total carbon emission limit ∑c i w i ≤Cmax, where c i For the first i Carbon emissions per unit of asset class, C max Maximum carbon emission limit; power grid transmission capacity constraint ∑P i w i ≤TTC k , where P i For the transmission power of asset class i, TTC k The upper limit of the transmission capacity of the k-th line; weight normalization condition w i ≥0 and ∑w i =1.

[0013] Furthermore, in step S3, the optimal weight scheme adopts the improved Sequential Quadratic Programming (SQP) algorithm, which decomposes the multi-asset portfolio problem into parallel computable subtasks. Combined with distributed reinforcement learning, the strategy is explored in a simulation environment through m virtual trading agents. The optimal solution is aggregated by the policy gradient algorithm, and the result is output.

[0014] Secondly, embodiments of this application also provide an energy asset portfolio optimization analysis system based on three-dimensional risk transmission, the system comprising: The data fusion module is used to collect multi-source heterogeneous data, including power data, meteorological data, industry standard text data and power equipment status data. It performs data cleaning and spatiotemporal alignment processing on the multi-source heterogeneous data, and then performs fusion processing on the processed multi-source heterogeneous data. a three-dimensional risk transmission model construction module, which constructs a three-dimensional risk transmission model for industry standard sensitivity, meteorological risk quantification, and power market fluctuation analysis; an optimization decision module, which constructs a multi-objective optimization framework according to the output of the three-dimensional risk transmission model, and generates an optimal weight scheme of the energy asset investment portfolio by maximizing the difference between expected returns and risk adjustment items, in combination with constraint conditions.

[0015] In a third aspect, the embodiments of the present application further provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the energy asset investment portfolio optimization analysis method based on three-dimensional risk transmission according to the first aspect when executing the program.

[0016] In a fourth aspect, the embodiments of the present application further provide a storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the steps of the energy asset investment portfolio optimization analysis method based on three-dimensional risk transmission according to the first aspect.

[0017] As can be seen from the above technical solutions, the present application has the following advantages: In the energy asset investment portfolio optimization analysis method, system, device and medium based on three-dimensional risk transmission according to the present application, multi-source heterogeneous data is collected and cleaned, time-space alignment and fusion are performed, the isolated forest algorithm and the Apache Kafka stream processing platform are used, the data quality and consistency are effectively improved, the traditional data silos are broken, a reliable and standardized time-space data foundation is provided for subsequent analysis, and the availability and accuracy of the data are ensured; the three-dimensional risk transmission model is constructed, advanced algorithms such as BERT, U-Net and improved GARCH-M are used, risk factors are accurately quantified from the dimensions of industry standards, meteorology and power market, the industry standard sensitivity analysis realizes semantic depth analysis of text, the meteorological risk quantification can accurately predict the availability decay of equipment affected by disasters, and the power market fluctuation coupling dynamically captures the linkage risk of the market and the equipment, so that the complex risks faced by energy assets are described from multiple dimensions and in all directions, and the risk assessment is more accurate; the multi-objective optimization framework combines the constraint conditions of carbon emission and power grid transmission, and uses the improved SQP algorithm and distributed reinforcement learning to maximize the returns while taking into account risk control and various real constraints, thereby efficiently generating an optimal asset weight scheme.

[0018] Not only does it balance returns and risks in energy asset allocation, but it also meets the requirements of low-carbon development and stable power grid, provides strong support for scientific investment decisions and resource optimization in the energy field, and promotes the efficient and sustainable development of the energy industry in complex environments. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the present application, the drawings required to be used in the description will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0020] Figure 1 The flowchart of the energy asset portfolio optimization analysis method based on three-dimensional risk transmission of the present application.

[0021] Figure 2 The schematic diagram of the energy asset portfolio optimization analysis system based on three-dimensional risk transmission of the present application. DETAILED DESCRIPTION

[0022] In the following detailed description of the specific steps of the energy asset portfolio optimization analysis method based on three-dimensional risk transmission, various embodiments of the present disclosure will be described more fully. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives falling within the spirit and scope of various embodiments of the present disclosure.

[0023] It should be understood that when used in the specification of the present application, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or sets thereof. The terms "comprise", "include", "have" and their variants mean "including but not limited to", unless otherwise specifically emphasized.

[0024] The phrase "one embodiment" or "some embodiments" or similar phrases as used in the present application means that a particular feature, structure or characteristic described in the embodiment is included in one or more embodiments of the present application. Therefore, the phrases "in one embodiment", "in some embodiments", "in other some embodiments", "in further some embodiments" and the like appearing in different places in the present application do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized.

[0025] In order to make the application purpose, features and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0026] Please refer to Figure 1 Fig. 1 shows a flowchart of an energy asset portfolio optimization analysis method based on three-dimensional risk transmission in a specific embodiment, including the following steps: S1. Collecting multi-source heterogeneous data including power data, meteorological data, industry standard text data and power equipment state data, performing data cleaning, space-time alignment processing on the multi-source heterogeneous data, and performing fusion processing on the processed multi-source heterogeneous data; It should be noted that by collecting multi-source heterogeneous data of power, meteorology, industry standard text and equipment state, combining isolated forest algorithm to clean abnormal data, Apache Kafka to realize space-time alignment, and then forming standardized space-time data through fusion, the data format difference and quality hidden danger can be eliminated, the multi-source data fragmentation problem can be solved, a unified and accurate data basis is provided for subsequent three-dimensional risk transmission modeling, the reliability of risk analysis and investment optimization is greatly improved, and decision-making errors caused by data deviation are avoided; S2. Constructing a three-dimensional risk transmission model for industry standard sensitivity, meteorological risk quantification and power market fluctuation analysis; It should be noted that by constructing a three-dimensional risk transmission model containing industry standards, meteorological risks and power market fluctuations, systematic disassembly and coupling analysis of multiple risk factors are realized, the limitations of traditional single-dimensional risk assessment are broken through, the dynamic influence of each factor on assets is accurately quantified, reliable risk factors are generated, scientific basis is provided for investment portfolio risk measurement, and the comprehensiveness and accuracy of risk identification are effectively improved; S3. According to the output result of the three-dimensional risk transmission model, a multi-objective optimization framework is constructed to maximize the difference between expected return and risk adjustment term, and the optimal weight scheme of energy asset portfolio is generated combined with constraint conditions; It should be noted that based on the three-dimensional risk model result, a multi-objective optimization framework is constructed to maximize the difference between return and risk adjustment term, combined with carbon emission and transmission capacity constraints, and the optimal weight scheme is generated through improved algorithm. It not only guarantees maximum return, but also meets the compliance requirements, solves the problem of ignoring multi-dimensional constraints in traditional optimization, improves the scientificity and feasibility of investment decision-making, and realizes the global optimal allocation of asset portfolio.

[0027] Further, as a refinement and extension of the above embodiment, in order to complete the specific implementation process in the embodiment, another energy asset portfolio optimization analysis method based on three-dimensional risk transmission is provided, which comprises: S1. Collect multi-source heterogeneous data, including power data, meteorological data, industry standard text data and power equipment state data, perform data cleaning, space-time alignment processing on the multi-source heterogeneous data, and perform fusion processing on the processed multi-source heterogeneous data; in step S1: The power data includes node price and transaction information of the power market, the meteorological data includes historical disaster satellite image data, and the power equipment state data includes SCADA data and life model data of device data acquisition and monitoring control system; The data cleaning uses the isolation forest algorithm to filter abnormal data, and the space-time alignment uses the Apache Kafka stream processing platform to align the time stamp; In step S1, the fusion processing of the processed multi-source heterogeneous data comprises: integrating the multi-source data after data cleaning and space-time alignment into standardized space-time data through unified data format and structure; Exemplarily, the node price of the power market, the day-ahead / real-time transaction record and other power data are collected; the historical disaster image data of typhoon, sandstorm and the like obtained by the meteorological satellite in the past 5 years are synchronized; the published new energy grid-connected power generation technology standards, carbon emission standards and other text data; and the device state data collected by the wind farm SCADA system, such as fan speed, photovoltaic inverter efficiency and equipment life prediction model data based on Weibull distribution; in the data cleaning stage, the isolation forest algorithm is used to detect and filter abnormal values (such as SCADA jump data caused by sensor failure, and price monitoring error values); by deploying the Apache Kafka stream processing platform, the multi-source data is uniformly mapped to the UTC time stamp, the time axis of the power transaction data, the meteorological image and the equipment state is synchronized, and the spatial grid is divided according to the region, and the spatial dimension is aligned; the cleaned and aligned data is converted into JSON format, and a standardized space-time database containing time, space and index is constructed, wherein each record contains a time stamp, a latitude and longitude coordinate, a data type label and a corresponding index value, providing a standardized input for the subsequent three-dimensional risk transmission model; S2. Construct a three-dimensional risk transmission model for industry standard sensitivity, meteorological risk quantification and power market fluctuation analysis; the specific steps of step S2 are: S21. The BERT model is used to perform semantic parsing on industry standard text data, extract key constraint factors including technical specifications, environmental protection requirements, and safety standards, and construct the influence function of industry standard factors by establishing the mapping relationship between constraint factors and asset return rate, and output the correction coefficient of asset return rate. S22. Use historical disaster satellite image data to train U-Net convolutional neural network, construct prediction models for wind power icing damage and photovoltaic dust shading, obtain 1km resolution wind power icing level probability map and photovoltaic dust shading probability map of power equipment through prediction model, map wind power icing level probability map and photovoltaic dust shading probability map to the influence function of meteorological risk factors, and output availability attenuation coefficient. S23. Using an improved GARCH-M model, conditional volatility is incorporated into the mean equation. Power data and power equipment data are input into the improved GARCH-M model, and the conditional mean and conditional volatility are output. An impact function of power market volatility factors is constructed, and the asset risk correction coefficient is output. S24. Couple the influence functions of industry standard factors, meteorological risk factors, and electricity market volatility factors to calculate the dynamic risk factor. Use the conditional value of risk (CVaR+) model to calculate the value of risk (CVaR) at confidence level α every 5 minutes. α :

[0028] Among them, R i For dynamic risk factors, γ is the value at risk at confidence level α, and p i Let be the probability of occurrence of the i-th asset class.

[0029] In step S24, the dynamic risk factor R i The calculation formula is:

[0030] Among them, R 0,i For the benchmark return, β i θ i φ i These represent the impact weights of industry standards, weather risks, and electricity market volatility, used to measure the degree of impact of different factors on asset returns; S i M is the output of the influence function of industry standard factors. i E is the output of the influence function of meteorological risk factors. i The output of the function representing the impact of electricity market fluctuations; For example, in the construction of the three-dimensional risk transmission model, the industry standard text is first preprocessed, and semantic features are extracted by word segmentation using the BERT model to identify constraint factors such as equipment efficiency standards in technical specifications, carbon emission limits in environmental protection requirements, and protection levels in safety standards; then, the mapping relationship between these factors and asset returns is established through multiple regression analysis to form an influence function, and the function outputs the corresponding reduced return correction coefficient. Historical satellite images of typhoons, sandstorms, and other disasters are correlated with concurrent damage data for wind power and photovoltaic equipment. This data is then input into a U-Net network for training, enabling the model to learn the correlation between image texture features and equipment damage. The trained model analyzes real-time satellite images to generate probability maps of wind power icing levels (levels 1-5) and photovoltaic dust shading (0-100%) across a 1km grid. Based on the probability distribution, the availability attenuation coefficient of equipment in each grid is calculated. The improved GARCH-M model incorporates equipment failure rate as an exogenous variable, inputting electricity price, load, and other power data, as well as equipment maintenance records. After model fitting, it outputs conditional mean (reflecting expected asset returns) and conditional volatility (reflecting market volatility). An influence function related to conditional mean and conditional volatility is constructed to obtain the asset risk correction coefficient. β was determined using the analytic hierarchy process. i θ i φ i Weights, S i M i E i Multiply each product by its respective weight and sum them to the benchmark return R. 0,i The dynamic risk factor Ri is obtained; every 5 minutes based on R... i The probability of asset occurrence p i The CVaR+ model is used to calculate the value of risk at a 95% confidence level, quantifying the potential loss under extreme risks. S3. Based on the output of the three-dimensional risk transmission model, construct a multi-objective optimization framework with the goal of maximizing the difference between expected return and risk adjustment term, and generate the optimal weight scheme for the energy asset portfolio in combination with constraints.

[0031] In step S3, the objective function of the multi-objective optimization framework is:

[0032] Among them, w i Let E(R) be the weight of the i-th asset class. i ) represents the expected return, λ represents the risk aversion coefficient, and CVaR α Value at risk at confidence level α; The constraints include: a total carbon emission limit ∑c i w i≤Cmax, where c i For the first i Carbon emissions per unit of asset class, C max Maximum carbon emission limit; power grid transmission capacity constraint ∑P i w i ≤TTC k , where P i For the transmission power of asset class i, TTC k The upper limit of the transmission capacity of the k-th line; weight normalization condition w i ≥0 and ∑w i =1.

[0033] In step S3, the optimal weight scheme adopts the improved Sequential Quadratic Programming (SQP) algorithm, which decomposes the multi-asset portfolio problem into parallel computable subtasks. Combined with distributed reinforcement learning, the strategy is explored in the simulation environment through m virtual trading agents. The optimal solution is aggregated by the policy gradient algorithm and the result is output. For example, in constructing a multi-objective optimization framework, the optimization objects are first clearly defined as four types of energy assets: wind power, photovoltaic power, hydropower, and thermal power, based on the dynamic risk factor R. i Calculate the expected return E(R) of various assets. i ), and set a risk aversion coefficient λ=0.6 to reflect a moderate risk preference; Transform the objective function into a concrete expression: max∑[w i ·E(R i -0.6w i ·CVaRα,i], where CVaRα,i is the 95% confidence level value of risk for each type of asset; When the constraints are implemented, the carbon emission constraints are set at c4 = 0.8 t / MWh for thermal power plants and c1 = 0.02 t / MWh for wind power plants, with a total limit C. max =5000t; Grid transmission capacity constraints apply to three lines: TTC1=1000MW, TTC2=800MW, and TTC3=600MW, corresponding to the allocated transmission power P for various assets. i (e.g., photovoltaic P2 = 300MW).

[0034] The improved SQP algorithm decomposes the four asset classes into two parallel subtasks (wind and solar combination, and water and fire combination). Ten virtual trading agents are launched to explore weight strategies in a simulated market environment. The agents try and fail through reinforcement learning, and the optimal solution is aggregated through the policy gradient algorithm. Finally, the weight scheme is output, which satisfies all constraints and maximizes the objective function value.

[0035] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0036] Please see Figure 2 As shown, the following are embodiments of the energy asset portfolio optimization analysis system based on three-dimensional risk transmission provided in this application. This system and the energy asset portfolio optimization analysis method based on three-dimensional risk transmission in the above embodiments belong to the same inventive concept. For details not described in detail in the embodiments of the energy asset portfolio optimization analysis system based on three-dimensional risk transmission, please refer to the embodiments of the energy asset portfolio optimization analysis method based on three-dimensional risk transmission described above.

[0037] The system includes: The data fusion module is used to collect multi-source heterogeneous data, including power data, meteorological data, industry standard text data and power equipment status data. It performs data cleaning and spatiotemporal alignment processing on the multi-source heterogeneous data, and then performs fusion processing on the processed multi-source heterogeneous data. The 3D risk transmission model construction module is used to build a 3D risk transmission model for industry standard sensitivity, meteorological risk quantification, and power market volatility analysis. The optimization decision module is used to construct a multi-objective optimization framework based on the output of the three-dimensional risk transmission model. The goal is to maximize the difference between the expected return and the risk adjustment term, and to generate the optimal weight scheme for the energy asset portfolio in combination with constraints.

[0038] This embodiment achieves low-threshold, high-efficiency investment decision-making and risk management through the collaboration of a data fusion module, a three-dimensional risk transmission model construction module, and an optimized decision-making module.

[0039] The energy asset portfolio optimization analysis method based on three-dimensional risk transmission provided in this application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiments of this invention does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, electronic devices include, but are not limited to, laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0040] Electronic devices may include processors, external memory interfaces, internal memory, universal serial bus (USB) interfaces, charging management modules, power management modules, batteries, wireless communication modules, audio modules, speakers, microphones, sensor modules, buttons, cameras, displays, and SIM card interfaces, etc.

[0041] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the electronic device. In other embodiments of this application, the electronic device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0042] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0043] The processor can serve as the nerve center and command center of an electronic device. The controller can generate operation control signals based on the instruction opcode and timing signals to control the fetching and execution of instructions.

[0044] The processor may also include memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or that are used repeatedly. If the processor needs to use the instruction or data again, it can retrieve it directly from this memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0045] The storage medium provided in this application stores a program product capable of implementing an energy asset portfolio optimization analysis method based on three-dimensional risk transmission.

[0046] In some possible implementations, the energy asset portfolio optimization analysis method based on three-dimensional risk transmission of this disclosure can be implemented as a program product, which includes program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure.

[0047] The storage medium disclosed herein can take the form of any combination of one or more readable media. A readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0048] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing energy asset portfolios based on three-dimensional risk transmission, characterized in that, Includes the following steps: S1. Collect multi-source heterogeneous data, including power data, meteorological data, industry standard text data and power equipment status data, perform data cleaning and spatiotemporal alignment processing on the multi-source heterogeneous data, and perform fusion processing on the processed multi-source heterogeneous data; S2. Construct a three-dimensional risk transmission model for analyzing industry standard sensitivity, meteorological risk quantification, and power market volatility. S3. Based on the output of the three-dimensional risk transmission model, construct a multi-objective optimization framework with the goal of maximizing the difference between expected return and risk adjustment term, and generate the optimal weight scheme for the energy asset portfolio in combination with constraints.

2. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 1, characterized in that, In step S1: The power data includes nodal electricity prices and transaction information in the power market; the meteorological data includes historical disaster satellite image data; and the power equipment status data includes SCADA data and life model data. The data cleaning uses the isolated forest algorithm to filter out abnormal data, and the spatiotemporal alignment uses the Apache Kafka streaming platform for timestamp alignment.

3. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 1, characterized in that, In step S1, the fusion processing of the processed multi-source heterogeneous data includes: integrating the cleaned and spatiotemporally aligned multi-source data into standardized spatiotemporal data through a unified data format and structure.

4. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 1, characterized in that, The specific steps of step S2 are as follows: S21. The BERT model is used to perform semantic parsing on industry standard text data, extract key constraint factors including technical specifications, environmental protection requirements, and safety standards, and construct the influence function of industry standard factors by establishing the mapping relationship between constraint factors and asset return rate, and output the correction coefficient of asset return rate. S22. Use historical disaster satellite image data to train U-Net convolutional neural network, construct prediction models for wind power icing damage and photovoltaic dust shading, obtain 1km resolution wind power icing level probability map and photovoltaic dust shading probability map of power equipment through prediction model, map wind power icing level probability map and photovoltaic dust shading probability map to the influence function of meteorological risk factors, and output availability attenuation coefficient. S23. Using an improved GARCH-M model, conditional volatility is incorporated into the mean equation. Power data and power equipment data are input into the improved GARCH-M model, and the conditional mean and conditional volatility are output. An impact function of power market volatility factors is constructed, and the asset risk correction coefficient is output. S24. Couple the influence functions of industry standard factors, meteorological risk factors, and electricity market volatility factors to calculate the dynamic risk factor. Use the conditional value of risk (CVaR+) model to calculate the value of risk (CVaR) at confidence level α every 5 minutes. α : Among them, R i For dynamic risk factors, γ is the value at risk at confidence level α, and p i Let be the probability of occurrence of the i-th asset class.

5. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 4, characterized in that, In step S24, the dynamic risk factor R i The calculation formula is: Among them, R 0,i For the benchmark return, β i θ i φ i These represent the impact weights of industry standards, weather risks, and electricity market volatility, used to measure the degree of impact of different factors on asset returns; S i M is the output of the influence function of industry standard factors. i E is the output of the influence function of meteorological risk factors. i This is the output of the function representing the impact of electricity market fluctuations.

6. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 1, characterized in that, In step S3, the objective function of the multi-objective optimization framework is: Among them, w i Let E(R) be the weight of the i-th asset class. i ) represents the expected return, λ represents the risk aversion coefficient, and CVaR α Value at risk at confidence level α; The constraints include: a total carbon emission limit ∑c i w i ≤Cmax, where c i For the first i Carbon emissions per unit of asset class, C max Maximum carbon emission limit; power grid transmission capacity constraint ∑P i w i ≤TTC k , where P i For the transmission power of asset class i, TTC k The upper limit of the transmission capacity of the k-th line; weight normalization condition w i ≥0 and ∑w i =1.

7. The energy asset portfolio optimization analysis method based on three-dimensional risk transmission according to claim 1, characterized in that, In step S3, the optimal weight scheme adopts the improved Sequential Quadratic Programming (SQP) algorithm, which decomposes the multi-asset portfolio problem into parallel computable subtasks. Combined with distributed reinforcement learning, the strategy is explored in a simulation environment through m virtual trading agents. The optimal solution is aggregated by the policy gradient algorithm and the result is output.

8. A portfolio optimization analysis system for energy assets based on three-dimensional risk transmission, characterized in that, The system includes: The data fusion module is used to collect multi-source heterogeneous data, including power data, meteorological data, industry standard text data and power equipment status data. It performs data cleaning and spatiotemporal alignment processing on the multi-source heterogeneous data, and then performs fusion processing on the processed multi-source heterogeneous data. The three-dimensional risk transmission model construction module builds a three-dimensional risk transmission model for analysis of industry standard sensitivity, meteorological risk quantification, and power market volatility. The optimization decision module is used to construct a multi-objective optimization framework based on the output of the three-dimensional risk transmission model. The goal is to maximize the difference between the expected return and the risk adjustment term, and to generate the optimal weight scheme for the energy asset portfolio in combination with constraints.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the energy asset portfolio optimization analysis method based on three-dimensional risk transmission as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the energy asset portfolio optimization analysis method based on three-dimensional risk transmission as described in any one of claims 1 to 7.

Images