An electric power market energy storage data interaction feature analysis method and system

Abstract

The application belongs to the technical field of electric power data analysis, and discloses a power market energy storage data interaction feature analysis method and system, which comprises the following steps: collecting original data of a power market, energy storage equipment and environmental conditions, identifying and cleaning abnormal data in the original data through data preprocessing to generate a data set in a unified format; extracting relevant features of each type of original data in the data set, eliminating invalid features in the relevant features based on variance threshold analysis, and retaining effective features to form a feature set; based on a time series prediction algorithm, a power market prediction model is constructed to predict future demand data and power price fluctuations of the power market and evaluate market risks existing in the energy storage system. The application combines data collection, preprocessing, feature extraction, market prediction, optimization decision and dynamic feedback mechanism, and can effectively improve the scheduling accuracy and efficiency of the energy storage system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power data analysis, in particular to a power market energy storage data interaction feature analysis method and system. BACKGROUND

[0002] With the growing global energy demand and the concern for sustainable development, the operation of the electricity market and the dispatch of energy storage systems become particularly important, especially in the electricity market, where energy supply and demand fluctuate greatly and prices are unstable, which poses great challenges to the dispatch of the power system. Energy storage devices, as an important response measure, can store electrical energy when the electricity market price is low and release electrical energy when the electricity demand is high, thereby balancing the supply and demand of the electricity market and reducing the risk of fluctuations. However, how to efficiently and accurately dispatch energy storage devices to maximize their benefits and reduce energy costs remains a pressing problem.

[0003] Currently, although there are many dispatch methods based on data analysis in the electricity market, the data format difference between different data sources leads to inconsistent or incomplete information, affecting subsequent analysis and decision-making. Many prediction models cannot effectively capture the complexity of electricity market price fluctuations and demand changes, resulting in large prediction errors and affecting the accuracy of energy storage dispatch. Most energy storage dispatch systems rely on static rules or pre-set dispatch models and cannot dynamically adjust according to real-time data, leading to insufficient utilization of energy storage resources.

[0004] Therefore, how to provide a power market energy storage data interaction feature analysis method and system is a problem to be solved at present. SUMMARY

[0005] The embodiments of the present application provide a power market energy storage data interaction feature analysis method and system to solve the above technical problems existing in the prior art.

[0006] To have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not a general review, nor is it intended to determine key / important components or delineate the scope of protection of these embodiments. Its only purpose is to present some concepts in a simple form as a prelude to the detailed description that follows.

[0007] According to a first aspect of the embodiments of the present application, a power market energy storage data interaction feature analysis method is provided.

[0008] In one embodiment, the power market energy storage data interaction feature analysis method comprises:

[0009] Collect raw data of a power market, energy storage devices, and environmental conditions, identify and clean abnormal data in the raw data through data preprocessing, and generate a data set in a unified format;

[0010] Extract relevant features of each type of raw data in the data set, eliminate invalid features in the relevant features based on variance threshold analysis, and retain effective features to form a feature set;

[0011] Based on a time series prediction algorithm, a power market prediction model is constructed to predict future demand data and power price fluctuations in the power market and to assess market risks existing in the energy storage system;

[0012] An optimization objective and constraint condition of the power market are established, and a linear programming optimization algorithm is used to solve the optimal scheduling decision of the energy storage device;

[0013] Based on real-time data and changes in the power market, an online learning algorithm is used to optimize the scheduling decision of the energy storage device in real time, and the scheduling decision is continuously adjusted according to a feedback mechanism.

[0014] In one embodiment, the collecting raw data of a power market, energy storage devices, and environmental conditions, identifying and cleaning abnormal data in the raw data through data preprocessing, and generating a data set in a unified format comprises:

[0015] Raw data of a power market, energy storage devices, and environmental conditions are collected in real time, and the raw data includes power market data, energy storage device data, and environmental data;

[0016] The missing values in the raw data are processed using linear interpolation, redundant data are removed, and an abnormal threshold of the raw data is set to identify and clean abnormal data that does not meet the abnormal threshold;

[0017] The raw data is converted into standardized data with a mean of 0 and a standard deviation of 1, and the raw data from different sources is time-stamped and aligned to form a data set in a unified format.

[0018] In one embodiment, the extracting relevant features of each type of raw data in the data set, eliminating invalid features in the relevant features based on variance threshold analysis, and retaining effective features to form a feature set comprises:

[0019] The relevant features of the raw data are extracted, and the relevant features include time features, statistical features, historical data features, and difference features;

[0020] Obtaining power price and demand data of a current power market, performing Fourier transform on the power price and demand data, converting a time domain signal into a frequency domain signal by using a Fourier transform formula, extracting periodic components in different frequency domains, and analyzing frequency characteristics of the periodic components;

[0021] Calculating a variance of each of the related features, comparing the variance with a preset variance threshold, determining that the related feature is an invalid feature and being eliminated if the variance of the related feature is less than or equal to the preset variance threshold, and determining that the related feature is a valid feature and integrating all the valid features to form a feature set if the variance of the related feature is greater than the preset variance threshold.

[0022] In one embodiment, the time series prediction algorithm-based power market prediction model is constructed to predict future demand and power price fluctuation of the power market, and the market risk existing in the energy storage system is evaluated, including:

[0023] Obtaining power price and demand data of a current power market, determining orders of an autoregressive part and a moving average part based on an autoregressive integrated moving average model time prediction algorithm, and constructing a power market prediction model;

[0024] Dividing the preprocessed data set into a training set and a test set, training and optimizing the power market prediction model by using the training set and the test set respectively, and predicting future power market demand by using the trained power market prediction model to obtain a prediction result.

[0025] In one embodiment, an expression of the power market prediction model is:

[0026]

[0027] In the formula, Y t is power price and demand data of a current time; μ is a constant term, is a coefficient of an autoregressive (AR) part; θ j is a coefficient of a moving average (MA) part; ∈ t is an error term; Y t-i is an observation value at time t-i; p is an order of the autoregressive part; θ j is a coefficient of the moving average part; ∈ t-j is an error term at time t-j; and q is an order of the moving average part.

[0028] In one embodiment, the establishment of the optimization target and the constraint condition of the power market generates an optimal scheduling decision of the energy storage device based on the obtained real-time data and the prediction result, including:

[0029] The power price of the power market and the power power, the power price and the power quantity required to be purchased are acquired, and a target function of the power market is established with a cost minimization as a target;

[0030] Constraint conditions of the power market are set, a linear programming optimization algorithm is used to calculate a scheduling strategy of the energy storage device, and an optimal scheduling strategy of the energy storage device is output.

[0031] In an embodiment, the target function of the power market is:

[0032]

[0033] In the formula, C t is the power price; t represents a time index; P t is the power power required to be purchased from the power grid at time t; γ t is the power price of the energy storage device discharging at time t; E t is the power quantity of the energy storage device discharging at time t; and T represents the total time period length.

[0034] In an embodiment, the expression for calculating the scheduling strategy of the energy storage device by using the linear programming optimization algorithm is:

[0035] min f(P, E), subject-to, g(P, E) ≤ 0, h(P, E) = 0

[0036] In the formula, P is one of the decision variables; E is another decision variable; f(P, E) is the target function; g(P, E) is a function representing inequality constraints; and h(P, E) is a function representing equality constraints.

[0037] According to a second aspect of an embodiment of the present application, a power market energy storage data interaction feature analysis system is provided.

[0038] In an embodiment, the power market energy storage data interaction feature analysis system comprises:

[0039] A data acquisition module is configured to acquire original data of a power market, an energy storage device and environmental conditions, identify and clean abnormal data in the original data through data preprocessing, and generate a data set in a unified format;

[0040] An extraction modeling module is configured to extract relevant features of each type of original data in the data set, eliminate invalid features in the relevant features based on variance threshold analysis, and retain effective features to form a feature set;

[0041] A prediction analysis module is configured to construct a power market prediction model based on a time series prediction algorithm, predict future demand data and power price fluctuations of the power market, and evaluate market risks existing in an energy storage system.

[0042] a decision optimization module, configured to establish an optimization target and a constraint condition of the electricity market, and to solve an optimal scheduling decision of the energy storage device based on a linear programming optimization algorithm;

[0043] a feedback adjustment module, configured to optimize the scheduling decision of the energy storage device in real time based on real-time data and changes in the electricity market by using an online learning algorithm, and to continuously adjust the scheduling decision according to a feedback mechanism.

[0044] In an embodiment, the data collection module comprises a data acquisition module, a data preprocessing module and a standardized merging module, wherein,

[0045] the data acquisition module is configured to collect original data of the electricity market, the energy storage device and environmental conditions in real time, and the original data comprises electricity market data, energy storage device data and environmental data;

[0046] the data preprocessing module is configured to process missing values in the original data by using a linear interpolation method, remove redundant data, and set an abnormal threshold of the original data to identify and clean abnormal data that does not meet the abnormal threshold;

[0047] the standardized merging module is configured to convert the original data into standardized data with a mean of 0 and a standard deviation of 1, and to align the time stamps of the original data from different sources to form a data set in a unified format.

[0048] In an embodiment, the extraction modeling module comprises a feature extraction module, a Fourier transform module and a feature selection module, wherein,

[0049] the feature extraction module is configured to extract relevant features of the original data, and the relevant features comprise time features, statistical features, historical data features and differential features;

[0050] the Fourier transform module is configured to obtain power price and demand data of the current electricity market, perform Fourier transform on the power price and demand data, convert time domain signals into frequency domain signals by using a Fourier transform formula, extract periodic components in different frequency domains, and analyze frequency characteristics of the periodic components;

[0051] the feature selection module is configured to calculate variances of each of the relevant features, compare the variances with a preset variance threshold, determine that a relevant feature is an invalid feature and is removed if the variance of the relevant feature is less than or equal to the preset variance threshold, and determine that the relevant feature is an effective feature and integrate all effective features to form a feature set if the variance of the relevant feature is greater than the preset variance threshold.

[0052] In an embodiment, the prediction analysis module comprises a model establishment module and a training prediction module, wherein,

[0053] The model establishing module is configured to acquire power price and demand data of the power market at a current time, determine orders of an autoregressive part and a moving average part based on an autoregressive integrated moving average model time prediction algorithm, and construct a power market prediction model.

[0054] The training prediction module is configured to divide the preprocessed data set into a training set and a test set, train and optimize the power market prediction model by using the training set and the test set respectively, predict future power market demand by using the trained power market prediction model, and obtain a prediction result.

[0055] In one embodiment, an expression of the power market prediction model is as follows:

[0056]

[0057] In the expression, Y t represents power price and demand data of the power market at a current time; μ represents a constant term, represents a coefficient of an autoregressive (AR) part; θ j represents a coefficient of a moving average (MA) part; ∈ t represents an error term; Y t-i represents an observation value at time t-i; p represents an order of the autoregressive part; θ j represents a coefficient of the moving average part; ∈ t-j represents an error term at time t-j; and q represents an order of the moving average part.

[0058] In one embodiment, the decision optimization module includes a target setting module and a strategy output module, wherein,

[0059] The target setting module is configured to acquire power price, required purchased power, power price, and power quantity of the power market, and establish a target function of the power market with a cost minimization as a target.

[0060] The strategy output module is configured to set a constraint condition of the power market, calculate an energy storage device scheduling strategy by using a linear programming optimization algorithm, and output an optimal scheduling strategy of the energy storage device.

[0061] In one embodiment, the target function of the power market is as follows:

[0062]

[0063] In the expression, C t represents power price; t represents a time index; P t represents power required to be purchased from a power grid at time t; γ t represents a power price of discharging of the energy storage device at time t; and Et is the amount of energy discharged by the energy storage device at time t; T represents the total length of the time period.

[0064] In one embodiment, the expression for calculating the energy storage device scheduling strategy by using the linear programming optimization algorithm is:

[0065] min f(P, E), subject-to, g(P, E) <= 0, h(P, E) = 0

[0066] In the formula, P is one of the decision variables; E is another decision variable; f(P, E) is the objective function; g(P, E) is a function representing inequality constraints; and h(P, E) is a function representing equality constraints.

[0067] According to a third aspect of an embodiment of the present application, a computer device is provided.

[0068] In some embodiments, the computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0069] According to a fourth aspect of an embodiment of the present application, a computer readable storage medium is provided.

[0070] In one embodiment, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the above method.

[0071] The technical solution provided by the embodiments of the present application can have the following beneficial effects: The present application combines data acquisition, preprocessing, feature extraction, market prediction, optimization decision and dynamic feedback mechanism, which can effectively improve the scheduling accuracy and efficiency of the energy storage system, and the effects are embodied in the following aspects:

[0072] 1. By integrating real-time data from the power market, energy storage devices and the environment, and preprocessing and cleaning, data redundancy and missing data are effectively eliminated, ensuring high quality and consistency of the data, which can provide accurate and reliable input for subsequent modeling and analysis.

[0073] 2. By Fourier transform to extract the periodic characteristics of the power market price and demand, and to extract time series features, statistical features and historical data features from the preprocessed data, sufficient input information is provided for the prediction of the power market and the scheduling of the energy storage system. This feature extraction method can help the system to extract key features with high prediction value from complex power market data, improving the accuracy of the model.

[0074] 3. Based on ARIMA autoregressive integrated moving average model and other time series prediction algorithms, the demand and price fluctuations of the electricity market are predicted, and the market risk is evaluated by combining risk analysis methods. Through accurate market prediction, the system can provide guidance for the scheduling of energy storage devices in advance, reducing the risk brought by market fluctuations.

[0075] 4. Through linear programming optimization algorithm, the scheduling strategy of energy storage devices is dynamically adjusted according to real-time data and prediction results, ensuring that the charging and discharging operation of energy storage devices in different time periods is the most efficient, and maximizing the benefit or minimizing the cost. The system can also adjust the optimization strategy according to real-time feedback, further improving the dynamic adaptability of the system.

[0076] 5. Based on real-time data and market changes, online learning algorithm is used to optimize the scheduling strategy of energy storage system, which enables the system to automatically adjust the operation strategy of energy storage devices when the electricity market and environmental conditions change, ensuring the flexibility and response speed of the system.

[0077] 6. Through accurate prediction based on market price and demand fluctuation, combined with real-time data adjustment of scheduling strategy, energy storage devices can more efficiently charge at low price and discharge at high price, maximizing the economic benefit and market benefit of energy storage devices.

[0078] 7. The optimized scheduling strategy can reduce unnecessary energy waste and excessive consumption, reduce the operating cost of the energy storage system, at the same time, accurate market prediction and risk analysis can effectively avoid the risk brought by market fluctuations, helping decision makers make more favorable decisions in uncertain market environment.

[0079] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0080] The accompanying drawings, which are incorporated into the specification and constitute part of it, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the application.

[0081] Figure 1 is a flow chart of a power market energy storage data interaction feature analysis method according to an exemplary embodiment;

[0082] Figure 2 is a structural schematic diagram of a power market energy storage data interaction feature analysis system according to an exemplary embodiment;

[0083] Figure 3 is a structural schematic diagram of a computer device according to an exemplary embodiment. DETAILED DESCRIPTION

[0084] The following description and drawings are illustrative of specific embodiments thereof and are not intended to limit the scope of the embodiments. Parts and features of some embodiments can be included or substituted in or for parts and features of other embodiments. The scope of the embodiments encompassed herein includes the whole scope of the claims together with all available equivalents of the claims. In this document, the terms "first", "second", etc. are used merely to distinguish one element from another, and do not require or imply any actual relationship or order between the elements. In fact, the first element can be referred to as the second element, and vice versa. Also, the terms "comprises", "comprising", or any other variations thereof are intended to cover a non-exclusive inclusion, such that a structure, device, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such structure, device, or apparatus. Without further limitation, an element defined by an "includes a" statement does not exclude the presence of additional identical elements in the structure, device, or apparatus that includes the element. Various embodiments are described in progressive stages, each of which focuses on the differences from other embodiments, and the same or similar parts between various embodiments can be referred to each other.

[0085] The terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like, as used herein, indicate relative positions or orientation relationships based on the positions or orientation relationships shown in the drawings, and are only used for the convenience of description herein and simplification of description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In the description herein, unless otherwise specified and limited, the terms "mount", "connect", "connection" should be understood broadly, for example, it can be a mechanical connection or an electrical connection, it can be a communication between two elements inside, it can be a direct connection, or an indirect connection through an intermediate medium, and the specific meaning of the above terms can be understood by those skilled in the art according to the specific circumstances.

[0086] In this document, the term "plurality" means two or more, unless otherwise specified.

[0087] In this document, the character " / " represents an "or" relationship between the objects before and after it. For example, A / B means A or B.

[0088] In this document, the term "and / or" is a description of the relationship between the objects, which means that there can be three relationships. For example, A and / or B means that there are three relationships of A or B, or A and B.

[0089] It should be understood that although the steps in the flowchart are shown in a sequential order, the steps are not necessarily performed in the order shown by the arrows. Unless explicitly stated otherwise, the steps can be performed in other orders. Moreover, at least some of the steps can include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but can be performed at different times, and the order of the sub-steps or stages can not be sequential, but can be interleaved or alternated with at least some of the other steps or sub-steps or stages of other steps.

[0090] The various modules in the device or system of the present application can be implemented in whole or in part by software, hardware, and combinations thereof. The various modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the various modules.

[0091] The embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0092] Figure 1 An embodiment of a power market energy storage data interaction feature analysis method of the present application is shown.

[0093] In this optional embodiment, the power market energy storage data interaction feature analysis method comprises:

[0094] Step S101, collect raw data of the power market, energy storage devices, and environmental conditions, identify and clean up abnormal data in the raw data through data preprocessing, and generate a data set in a unified format;

[0095] Step S103, extract relevant features of each type of raw data in the data set, eliminate invalid features in the relevant features based on variance threshold analysis, and retain effective features to form a feature set;

[0096] Step S105, based on a time series prediction algorithm, construct a power market prediction model, predict future demand data and power price fluctuations of the power market, and evaluate market risks existing in the energy storage system;

[0097] Step S107, establish an optimization target and constraint condition of the power market, and solve the optimal scheduling decision of the energy storage device based on a linear programming optimization algorithm;

[0098] Step S109, based on real-time data and changes in the power market, use an online learning algorithm to optimize the scheduling decision of the energy storage device in real time, and continuously adjust the scheduling decision according to a feedback mechanism.

[0099] In the optional embodiment, when the raw data of the electricity market, energy storage device and environmental conditions is collected, the raw data is preprocessed, the abnormal data in the raw data is identified and cleaned, and the data set in a unified format is generated, the raw data of the electricity market, energy storage device and environmental conditions can be collected in real time, and the raw data includes electricity market data, energy storage device data and environmental data; the missing values in the raw data are processed by using linear interpolation method, redundant data is removed, and an abnormal threshold of the raw data is set to identify and clean abnormal data that does not meet the abnormal threshold; the raw data is converted into standardized data with a mean value of 0 and a standard deviation of 1, and the raw data from different sources is time-stamped and aligned to form a data set in a unified format.

[0100] In the optional embodiment, when the relevant features of each type of raw data in the data set are extracted, the invalid features in the relevant features are removed based on variance threshold analysis, and the effective features are retained to form a feature set, the relevant features of the raw data can be extracted, and the relevant features include time features, statistical features, historical data features and difference features; the electricity price and demand data of the current electricity market are obtained, the electricity price and demand data are subjected to Fourier transform, the time domain signal is converted into a frequency domain signal by using Fourier transform formula, the periodic components in different frequency domains are extracted, and the frequency characteristics of the periodic components are analyzed; the variance of each relevant feature is calculated, and compared with the preset variance threshold; if the variance of the relevant feature is less than or equal to the preset variance threshold, it is determined that the relevant feature is invalid and is removed; if the variance of the relevant feature is greater than the preset variance threshold, it is determined that the relevant feature is effective, and all effective features are integrated to form a feature set.

[0101] In the optional embodiment, when the electricity market prediction model is constructed based on the time series prediction algorithm, the future demand and electricity price fluctuation of the electricity market are predicted, and the market risk existing in the energy storage system is evaluated, the electricity price and demand data of the current electricity market are obtained, the order of the autoregressive part and the moving average part is determined based on the autoregressive integrated moving average model time prediction algorithm, and the electricity market prediction model is constructed; the preprocessed data set is divided into a training set and a test set, the training set and the test set are used to train and optimize the electricity market prediction model respectively, the trained electricity market prediction model is used to predict the future electricity market demand, and the prediction result is obtained.

[0102] In the optional embodiment, the expression of the electricity market prediction model is:

[0103]

[0104] In the formula, Y tis the power market price and demand data at the current time; μ is a constant term, is the coefficient of the autoregressive (AR) part; θ j is the coefficient of the moving average (MA) part; ∈ t is the error term; Y t-i is the observation value at time t-i; p is the order of the autoregressive part; θ j is the coefficient of the moving average part; ∈ t-j is the error term at time t-j; q is the order of the moving average part.

[0105] In this optional embodiment, when the optimization target and constraint condition of the power market are established, and the optimal scheduling decision of the energy storage device is generated based on the obtained real-time data and prediction results, the power price and the required purchased power, the electricity price and the electricity quantity of the power market are obtained, and the objective function of the power market is established with the target of minimizing the cost; the constraint condition of the power market is set, the linear programming optimization algorithm is used to calculate the scheduling strategy of the energy storage device, and the optimal scheduling strategy of the energy storage device is output.

[0106] In this optional embodiment, the objective function of the power market is:

[0107]

[0108] In the formula, C t is the power price; t represents the time index; P t is the power required to be purchased from the power grid at time t; γ t is the electricity price of the energy storage device discharging at time t; E t is the electricity quantity of the energy storage device discharging at time t; T represents the total time period length.

[0109] In this optional embodiment, the expression for calculating the scheduling strategy of the energy storage device by using the linear programming optimization algorithm is:

[0110] minf(P,E),subject-to,g(P,E)≤0,h(P,E)=0

[0111] In the formula, P is one of the decision variables; E is another decision variable; f(P,E) is the objective function; g(P,E) is a function representing the inequality constraint; h(P,E) is a function representing the equality constraint.

[0112] Figure 2 An embodiment of the power market energy storage data interaction feature analysis system of the application is shown.

[0113] In this optional embodiment, the power market energy storage data interaction feature analysis system comprises:

[0114] The data collection module 201 is configured to collect raw data of the electricity market, the energy storage device and the environmental condition, to identify and clean abnormal data in the raw data through data preprocessing, and to generate a data set in a unified format.

[0115] The feature extraction and modeling module 203 is configured to extract relevant features of each type of raw data in the data set, to remove invalid features in the relevant features based on variance threshold analysis, and to retain valid features to form a feature set.

[0116] The prediction analysis module 205 is configured to construct an electricity market prediction model based on a time series prediction algorithm, to predict future demand data and electricity price fluctuations of the electricity market, and to evaluate market risks existing in the energy storage system.

[0117] The decision optimization module 207 is configured to establish an optimization target and constraint condition of the electricity market, and to solve an optimal scheduling decision of the energy storage device based on a linear programming optimization algorithm.

[0118] The feedback adjustment module 209 is configured to use an online learning algorithm to optimize the scheduling decision of the energy storage device in real time based on real-time data and changes in the electricity market, and to continuously adjust the scheduling decision according to a feedback mechanism.

[0119] In the optional embodiment, the data collection module 201 includes a data acquisition module (not shown in the figure), a data preprocessing module (not shown in the figure) and a standardization and merging module (not shown in the figure). The data acquisition module is configured to collect raw data of the electricity market, the energy storage device and the environmental condition in real time, and the raw data includes electricity market data, energy storage device data and environmental data. The data preprocessing module is configured to process missing values in the raw data by using a linear interpolation method, to remove redundant data, to set an abnormal threshold of the raw data, and to identify and clean abnormal data that does not meet the abnormal threshold. The standardization and merging module is configured to convert the raw data into standardized data with a mean value of 0 and a standard deviation of 1, and to perform time stamp alignment on the raw data from different sources to form a data set in a unified format.

[0120] In the optional embodiment, the feature extraction modeling module 203 comprises a feature extraction module (not shown in the figure), a Fourier transform module (not shown in the figure), and a feature selection module (not shown in the figure). The feature extraction module is configured to extract relevant features of the original data, and the relevant features include time features, statistical features, historical data features, and differential features. The Fourier transform module is configured to obtain power price and demand data of a current power market, perform Fourier transform on the power price and demand data, convert time domain signals into frequency domain signals by using a Fourier transform formula, extract periodic components in different frequency domains, and analyze frequency features of the periodic components. The feature selection module is configured to calculate variances of each of the relevant features, compare the variances with a preset variance threshold, determine that a relevant feature is an invalid feature and is removed if the variance of the relevant feature is less than or equal to the preset variance threshold, and determine that the relevant feature is a valid feature and integrate all valid features to form a feature set if the variance of the relevant feature is greater than the preset variance threshold.

[0121] In the optional embodiment, the prediction analysis module 205 comprises a model establishment module (not shown in the figure) and a training prediction module (not shown in the figure). The model establishment module is configured to obtain power price and demand data of a current power market, determine orders of an autoregressive part and a moving average part based on an autoregressive integrated moving average model time prediction algorithm, and construct a power market prediction model. The training prediction module is configured to divide a preprocessed data set into a training set and a test set, train and optimize the power market prediction model by using the training set and the test set respectively, predict future power market demand by using the trained power market prediction model, and obtain a prediction result.

[0122] In the optional embodiment, an expression of the power market prediction model is as follows:

[0123]

[0124] wherein Y t is power market price and demand data at a current time; μ is a constant term, is a coefficient of an autoregressive (AR) part; θ j is a coefficient of a moving average (MA) part; ∈ t is an error term; Y t-i is an observation value at time t-i; p is an order of the autoregressive part; θ j is a coefficient of the moving average part; ∈ t-j is an error term at time t-j; and q is an order of the moving average part.

[0125] In the optional embodiment, the decision optimization module 207 comprises a target setting module (not shown in the figure) and a strategy output module (not shown in the figure), wherein the target setting module is configured to obtain the electricity price of the electricity market and the required purchased electricity power, electricity price and electricity quantity, and to establish a target function of the electricity market with the cost minimization as the target; and the strategy output module is configured to set the constraint condition of the electricity market, to calculate the energy storage device scheduling strategy by using a linear programming optimization algorithm, and to output the optimal scheduling strategy of the energy storage device.

[0126] In the optional embodiment, the target function of the electricity market is as follows:

[0127]

[0128] wherein C t is the electricity price; t represents the time index; P t is the electricity power required to be purchased from the power grid at time t; γ t is the electricity price of the energy storage device discharging at time t; E t is the electricity quantity of the energy storage device discharging at time t; and T represents the total time period length.

[0129] In the optional embodiment, the expression for calculating the energy storage device scheduling strategy by using the linear programming optimization algorithm is as follows:

[0130] min f(P, E), subject-to, g(P, E)≤0, h(P, E)=0

[0131] wherein P is one of the decision variables; E is another decision variable; f(P, E) is the target function; g(P, E) is a function representing the inequality constraint; and h(P, E) is a function representing the equality constraint.

[0132] The application will be further described in detail below in combination with the drawings and specific embodiments.

[0133] A power market energy storage data interaction feature analysis system comprises a data acquisition module 201, an extraction modeling module 203, a prediction analysis module 205, a decision optimization module 207 and a feedback adjustment module 209.

[0134] 1. For the data acquisition module 201, various types of power market, energy storage device and environmental condition data are acquired, data of different sources are integrated, the collected original data are preprocessed, and the abnormal data are automatically identified and cleaned based on an abnormality detection algorithm. The composition structure and implementation principle of the data acquisition module 201 include the following aspects:

[0135] 1.1 Data acquisition module: real-time acquisition of power market data, energy storage device data and environmental data through sensors.

[0136] 1.2 Data preprocessing module: remove redundant data: remove duplicate records, use linear interpolation algorithm for missing value processing, where the linear interpolation algorithm formula is:

[0137]

[0138] In the formula, x(t) is the missing value, t0 and t1 are the time points before and after, x(t0) and x(t1) are the known data at the corresponding time;

[0139] Based on statistical anomaly detection, use statistical method to detect data points that do not conform to the expected distribution, if the data point exceeds the predetermined threshold, it is considered that the data is abnormal, wherein the anomaly detection formula is:

[0140]

[0141] In the formula, x is the data point, μ is the mean of the data, σ is the standard deviation of the data, and z is a standardized index that measures how far the data point is from the mean;

[0142] 1.3 Standardization and merging module: standardize the data, transform the data into a distribution with mean 0 and standard deviation 1; Align the timestamps of data from different sources, and fuse information from different data sources to generate a set of unified format data sets.

[0143] 2, for the extraction modeling module 203, electrically connected with the data acquisition module 201, the extraction modeling module 203 is used to extract the characteristics of power energy storage scheduling and market analysis from the preprocessed data. The composition structure and implementation principle of the extraction modeling module 203 include the following aspects:

[0144] 2.1 Feature extraction module: extract features from preprocessed data, including time features, statistical features, historical data features and difference features.

[0145] 2.2 Fourier transform module: extract different frequency periodic components by performing Fourier transform on power price and demand data, wherein the Fourier transform formula is:

[0146]

[0147] In the formula, FT(f) represents the Fourier transform result at frequency f, x t represents the signal data in the time domain, N is the total length of the signal, f is the frequency, e -2πift / N is a complex exponential term, i is the imaginary unit, 2π is the periodic constant, and ft / N is a proportional factor.

[0148] 2.3 Feature selection module: calculate the variance of each feature, set a threshold, select the features with variance greater than the threshold, eliminate the features with variance close to zero, and construct a new feature set with the remaining features with larger variance.

[0149] 3. For the prediction analysis module 205, which is electrically connected with the extraction modeling module 203, the prediction analysis module 205 is used to predict the electricity market based on a time series prediction algorithm, including electricity price fluctuations and market demand, to provide forward-looking guidance for energy storage system scheduling. The composition structure and implementation principle of the prediction analysis module 205 include the following aspects:

[0150] 3.1 Model establishment module: based on the autoregressive integrated moving average model time series prediction algorithm, an electricity market prediction model is constructed, wherein the electricity market prediction model formula is:

[0151]

[0152] where Y t is the electricity market price and demand at the current time, μ is a constant term, is the coefficient of the AR part, θ j is the coefficient of the MA part, ∈ t is the error term, Y t-i is the observation value at time t-i, p is the order of the autoregressive part, θ j is the coefficient of the moving average part, ∈ t-j is the error term at time t-j, and q is the order of the moving average part.

[0153] 3.2 Training and prediction module: the preprocessed data is divided into training set and test set, the cross-validation method is used to train and verify the prediction model, and the trained model is used to predict the future electricity market demand.

[0154] 4. For the decision optimization module 207, which is electrically connected with the prediction analysis module 205, the decision optimization module 207 is used to provide scheduling strategies for energy storage devices based on real-time data and prediction data through optimization algorithms. The composition structure and implementation principle of the decision optimization module 207 include the following aspects:

[0155] 4.1 Objective setting module: based on cost minimization, an objective function is established, wherein the objective function formula is:

[0156]

[0157] where C t is the electricity price, t represents the time index, P t is the electricity power that needs to be purchased from the power grid at time t, and γ tis the electricity price at time t when the energy storage device discharges, E t is the electricity quantity at time t when the energy storage device discharges, T represents the total time period length.

[0158] 4.2 Strategy output module: set constraints to ensure that the scheduling strategy is executed within a feasible range, use a linear programming optimization algorithm to solve the energy storage scheduling problem; after solving by the optimization algorithm, the system outputs the optimal energy storage device scheduling strategy, combines these results with the real-time monitoring system, and adjusts the running state of the device in real time.

[0159] The use of a linear programming optimization algorithm to solve the energy storage scheduling problem specifically includes:

[0160] Wherein, the optimization formula is:

[0161] minf(P, E), subject-to, g(P, E)≤0, h(P, E)=0

[0162] In the formula, P represents one of the decision variables, E represents another decision variable, f(P, E) is the objective function, g(P, E) is a function representing inequality constraints, and h(P, E) is a function representing equality constraints.

[0163] 5、For the feedback adjustment module 209, which is electrically connected with the decision optimization module 207, the feedback adjustment module 209 is used to optimize the scheduling strategy of the energy storage system based on real-time data and market changes using an online learning algorithm.

[0164] In addition, an electricity market energy storage data interaction feature analysis can be summarized as follows:

[0165] Step one: Collect raw data of electricity market, energy storage device and environmental conditions through multiple data sources, including price fluctuations of electricity market, market demand, state of energy storage device and environmental condition data, clean and standardize the collected raw data, use an anomaly detection algorithm to automatically identify and clean abnormal data, and ensure the quality and consistency of the data.

[0166] Step two: Extract features related to electricity storage scheduling and market analysis from the preprocessed data, including time series features, electricity market features, energy storage device features and environmental condition features, and model analysis combined with interaction features.

[0167] Step three: Based on time series prediction algorithm, predict future demand and electricity price fluctuations of electricity market, and evaluate the market risk that may be faced in energy storage system scheduling by combining risk assessment method;

[0168] Step four: Use optimization algorithm to provide optimal scheduling strategy for energy storage device based on real-time data and prediction results.

[0169] Step five: Based on real-time data and market changes, online learning algorithms are used to optimize the scheduling strategy of the energy storage system, and the scheduling decision is continuously adjusted according to the feedback mechanism.

[0170] In an embodiment, a computer device, which can be a server, has an internal structure diagram as shown in Figure 3 The computer device includes a processor, a memory and a network interface connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store static information and dynamic information data. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement the steps in the above method embodiments.

[0171] Those skilled in the art can understand that Figure 3 The structure shown in the above

[0172] In addition, the present application also provides a computer device including a memory and a processor, the memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0173] In addition, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor to implement the steps in the above method embodiments.

[0174] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in each embodiment of the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0175] The present application is not limited to the structures described above and shown in the drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present application is only limited by the appended claims.

Claims

1. A method for analyzing the interactive characteristics of energy storage data in the electricity market, characterized in that, include: Raw data on the electricity market, energy storage devices, and environmental conditions are collected. Through data preprocessing, abnormal data in the raw data is identified and cleaned up, and a dataset in a unified format is generated. Extract relevant features from each type of raw data in the dataset, and based on variance threshold analysis, remove invalid features from the relevant features and retain valid features to form a feature set; Based on time series forecasting algorithms, an electricity market forecasting model is constructed to predict future demand data and electricity price fluctuations in the electricity market, and to assess the market risks existing in energy storage systems. Establish the optimization objectives and constraints of the electricity market, and solve the optimal scheduling decision for energy storage devices based on the linear programming optimization algorithm; Based on real-time data and changes in the electricity market, online learning algorithms are used to optimize the scheduling decisions of energy storage devices in real time, and the scheduling decisions are continuously adjusted according to the feedback mechanism. The step of extracting relevant features from the original data of each type in the dataset, and based on variance threshold analysis, removing invalid features from the relevant features and retaining valid features to form a feature set includes: Extract relevant features from the original data, including time features, statistical features, historical data features, and difference features; Obtain current electricity price and demand data in the electricity market, perform Fourier transform on the electricity price and demand data, use the Fourier transform formula to convert the time domain signal into a frequency domain signal, extract periodic components in different frequency domains, and analyze the frequency characteristics of the periodic components. Calculate the variance of each relevant feature and compare it with a preset variance threshold. If the variance of the relevant feature is less than or equal to the preset variance threshold, it is determined to be an invalid feature and is removed. If the variance of the relevant feature is greater than the preset variance threshold, it is determined to be a valid feature and all valid features are integrated to form a feature set. The optimization objectives and constraints for establishing the electricity market, based on the acquired real-time data and forecast results, generate the optimal scheduling decision for energy storage devices, including: To obtain the electricity price in the electricity market and the required power, price, and quantity of electricity to be purchased, and to establish the objective function of the electricity market with the goal of minimizing costs; Set constraints for the electricity market, use linear programming optimization algorithm to calculate the scheduling strategy for energy storage devices, and output the optimal scheduling strategy for energy storage devices.

2. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 1, characterized in that, The process of collecting raw data from the electricity market, energy storage devices, and environmental conditions, followed by data preprocessing to identify and clean up abnormal data, and generating a dataset in a unified format includes: Real-time collection of raw data on the electricity market, energy storage devices, and environmental conditions; The missing values ​​in the original data are processed using linear interpolation, redundant data is removed, and an anomaly threshold is set for the original data to identify and clean up abnormal data that does not meet the anomaly threshold. The original data is converted into standardized data with a mean of 0 and a standard deviation of 1, and the original data from different sources are timestamped and merged to form a dataset with a unified format.

3. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 1, characterized in that, The raw data includes electricity market data, energy storage equipment data, and environmental data.

4. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 1, characterized in that, The time-series forecasting algorithm is used to construct a power market forecasting model to predict future demand and price fluctuations in the power market, and to assess the market risks associated with energy storage systems, including: Obtain current electricity price and demand data in the electricity market, and construct an electricity market prediction model based on the time prediction algorithm of the autoregressive integral moving average model, determining the order of the autoregressive part and the moving average part. The preprocessed dataset is divided into a training set and a test set. The electricity market prediction model is trained and optimized using the training set and the test set, respectively. The trained electricity market prediction model is then used to predict future electricity market demand to obtain the prediction results.

5. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 4, characterized in that, The expression for the electricity market forecasting model is as follows: In the formula, Y t This represents the current electricity market price and demand data; μ is a constant term. The coefficients of the autoregressive component; θ j The coefficients of the moving average component; ∈ t Y is the error term; t-i θ represents the observed value at time ti; p is the order of the autoregressive component; θ j The coefficients of the moving average component; ∈ t-j q represents the error term for time tj; q is the order of the moving average component.

6. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 5, characterized in that, The objective function of the electricity market is: In the formula, C t For electricity price; t represents time index; P t γ is the amount of electricity that needs to be purchased from the grid at time t; t E is the electricity price at which the energy storage device discharges at time t; t It represents the amount of electricity discharged by the energy storage device at time t; T represents the total time period length.

7. The method for analyzing the interactive characteristics of energy storage data in the power market according to claim 6, characterized in that, The expression for calculating the energy storage device scheduling strategy using the linear programming optimization algorithm is as follows: minf(P,E),subject-to,g(P,E)≤0,h(P,E)=0 In the formula, P is one of the decision variables; E is the other decision variable; f(P,E) is the objective function; g(P,E) is the function representing inequality constraints; and h(P,E) is the function representing equality constraints.

8. A data interaction feature analysis system for energy storage in the electricity market, characterized in that, include: The data acquisition module is used to collect raw data on the electricity market, energy storage devices, and environmental conditions. Through data preprocessing, it identifies and cleans abnormal data in the raw data and generates a dataset in a unified format. The extraction and modeling module is used to extract relevant features of each type of raw data in the dataset, and based on variance threshold analysis, remove invalid features from the relevant features and retain valid features to form a feature set. The predictive analytics module is used to build electricity market forecasting models based on time series forecasting algorithms, predict future demand data and electricity price fluctuations in the electricity market, and assess market risks in energy storage systems. The decision optimization module is used to establish the optimization objectives and constraints of the electricity market, and solve the optimal scheduling decision of energy storage devices based on the linear programming optimization algorithm. The feedback adjustment module is used to optimize the scheduling decisions of energy storage devices in real time based on real-time data and changes in the electricity market, and to continuously adjust the scheduling decisions according to the feedback mechanism. The extraction and modeling module includes: a feature extraction module, a Fourier transform module, and a feature selection module, wherein, The feature extraction module is used to extract relevant features from the original data, and the relevant features include time features, statistical features, historical data features, and difference features. The Fourier transform module is used to acquire current electricity price and demand data in the electricity market, perform Fourier transform on the electricity price and demand data, convert the time domain signal into a frequency domain signal using the Fourier transform formula, extract periodic components in different frequency domains, and analyze the frequency characteristics of the periodic components. The feature selection module is used to calculate the variance of each relevant feature and compare it with a preset variance threshold. If the variance of the relevant feature is less than or equal to the preset variance threshold, it is determined to be an invalid feature and is removed. If the variance of the relevant feature is greater than the preset variance threshold, it is determined to be a valid feature and all valid features are integrated to form a feature set. The decision optimization module includes: a target setting module and a strategy output module, wherein... The target setting module is used to obtain the electricity price in the electricity market and the required power, price and quantity of electricity to be purchased, and to establish the objective function of the electricity market with the goal of minimizing costs. The strategy output module is used to set the constraints of the electricity market, use a linear programming optimization algorithm to calculate the energy storage device scheduling strategy, and output the optimal scheduling strategy for the energy storage device.

9. The power market energy storage data interaction characteristic analysis system according to claim 8, characterized in that, The data acquisition module includes: a data acquisition module, a data preprocessing module, and a standardization and merging module, wherein, The data acquisition module is used to collect raw data on the electricity market, energy storage devices, and environmental conditions in real time. The data preprocessing module is used to process missing values ​​in the original data using linear interpolation, remove redundant data, set an anomaly threshold for the original data, and identify and clean up abnormal data that does not meet the anomaly threshold. The standardization and merging module is used to convert the original data into standardized data with a mean of 0 and a standard deviation of 1, and to align the original data from different sources with timestamps and merge them into a dataset with a unified format.

10. The power market energy storage data interaction characteristic analysis system according to claim 8, characterized in that, The raw data includes electricity market data, energy storage equipment data, and environmental data.

11. The power market energy storage data interaction characteristic analysis system according to claim 8, characterized in that, The predictive analysis module includes: a model building module and a training prediction module, wherein... The model building module is used to obtain the electricity price and demand data of the electricity market at the current moment, and to determine the order of the autoregressive part and the moving average part based on the time prediction algorithm of the autoregressive integral moving average model, and to construct the electricity market prediction model. The training and prediction module is used to divide the preprocessed dataset into a training set and a test set, and to train and optimize the electricity market prediction model using the training set and the test set respectively. The trained electricity market prediction model is then used to predict future electricity market demand to obtain prediction results.

12. The power market energy storage data interaction characteristic analysis system according to claim 11, characterized in that, The expression for the electricity market forecasting model is as follows: In the formula, Y t This represents the current electricity market price and demand data; μ is a constant term. These are the coefficients of the autoregressive component; θ j The coefficients of the moving average component; ∈ t Y is the error term; t-i θ represents the observed value at time ti; p is the order of the autoregressive component; θ j The coefficients of the moving average component; ∈ t-j q represents the error term for time tj; q is the order of the moving average component.

13. The power market energy storage data interaction characteristic analysis system according to claim 12, characterized in that, The objective function of the electricity market is: In the formula, C t For electricity price; t represents time index; P t γ is the amount of electricity that needs to be purchased from the grid at time t; t E is the electricity price at which the energy storage device discharges at time t; t It represents the amount of electricity discharged by the energy storage device at time t; T represents the total time period length.

14. The power market energy storage data interaction characteristic analysis system according to claim 13, characterized in that, The expression for calculating the energy storage device scheduling strategy using the linear programming optimization algorithm is as follows: minf(P,E),subject-to,g(P,E)≤0,h(P,E)=0 In the formula, P is one of the decision variables; E is the other decision variable; f(P,E) is the objective function; g(P,E) is the function representing the inequality constraint; and h(P,E) is the function representing the equality constraint.

Images