Model data fusion-based equipment-level digital twinborn body construction method

Abstract

The invention provides an equipment-level digital twinborn body construction method based on model data fusion, and belongs to the technical field of renewable energy scheduling. The method comprises the following steps: collecting operation data and external environment data of photovoltaic power generation, wind power generation and energy storage equipment; performing data cleaning and data standardization processing on the operation data and the external environment data; constructing a digital twinborn body of the photovoltaic power generation equipment, and dynamically adjusting the digital twinborn body of the photovoltaic power generation equipment by using the external environment data, the operation data of the photovoltaic power generation equipment and the historical power generation data; constructing a digital twinborn body of the wind power generation equipment; and constructing a digital twinborn body of the energy storage equipment. The method has the advantages that through multi-source data fusion, refined modeling and dynamic optimization of equipment are realized, and accurate reflection and regulation and control of the equipment state are ensured; and through multi-source data fusion, refined modeling and dynamic optimization of the equipment are realized, and accurate reflection and regulation and control of the equipment state are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of renewable energy scheduling, and specifically relates to a device-level digital twin construction method based on model data fusion. Background Art

[0002] With the rapid development of renewable energy, photovoltaic power generation, wind power generation and energy storage equipment, as the most promising distributed energy resources in the power system, have been widely used worldwide. By giving full play to the flexibility of these distributed energy sources, the operating efficiency of the power system can be effectively improved and a higher proportion of clean energy consumption can be achieved. At the same time, compared with traditional centralized power sources, the construction cost of distributed energy resources is relatively low, and they can reduce carbon emissions, with significant economic and environmental benefits. However, distributed energy resources such as photovoltaic power generation and wind power generation are greatly affected by the external environment, and their output has large volatility and uncertainty. The charging and discharging strategy of energy storage equipment is constrained by load demand and grid scheduling. Therefore, how to build a digital twin that can accurately reflect the operating characteristics of these distributed energy devices and dynamically optimize them through real-time data is a key issue in the current technical field. Although the existing digital twin construction methods have made certain progress in large-scale energy systems, there are still many technical challenges for device-level modeling and optimization, especially in multivariate data fusion, real-time interaction and device adaptive control. In this context, developing a device-level digital twin construction method based on model and data fusion can effectively improve the dynamic perception capability of distributed energy resources and provide more accurate decision-making support for their scheduling optimization, which has important theoretical and practical significance.

[0003] Therefore, technicians in this field are committed to developing a device-level digital twin construction method based on model data fusion to achieve accurate modeling and optimized control of distributed energy equipment and improve the operating efficiency and stability of the system. Summary of the invention

[0004] The technical problem to be solved by the present invention is the scheduling optimization of distributed new energy. In view of the shortcomings of the existing technology, a device-level digital twin construction method based on model data fusion is provided, which can effectively improve the dynamic perception ability of distributed energy resources. In order to solve the above technical problems, the technical solution adopted by the present invention is as follows: collecting the operation data and external environment data of photovoltaic power generation, wind power generation and energy storage equipment;

[0005] Performing data cleaning and data standardization processing on the operation data and the external environment data;

[0006] Constructing a digital twin of the photovoltaic power generation equipment, and dynamically adjusting the digital twin of the photovoltaic power generation equipment using the external environment data, the operation data of the photovoltaic power generation equipment, and the historical power generation data;

[0007] Constructing a digital twin of the wind power generation equipment, and dynamically adjusting the digital twin of the wind power generation equipment using the external environment data, the operation data of the wind power generation equipment, and the historical power generation data;

[0008] Construct a digital twin of the energy storage device, and dynamically adjust the digital twin of the energy storage device using the external environment data, the operation data of the energy storage device, and the historical operation data

[0009] Furthermore, the method for collecting operation data and external environment data of photovoltaic power generation, wind power generation and energy storage equipment includes:

[0010] Real-time collection of operating data of photovoltaic power generation, wind power generation and energy storage equipment. The collected electrical quantity data include voltage, current and power output; the collected external environment data include temperature, light intensity and wind speed.

[0011] Furthermore, the method for cleaning the operation data and the external environment data includes:

[0012] The outliers in the data are removed by using the outlier detection algorithm. The outliers are detected by the 3σ principle. The mean and standard deviation of the collected data are analyzed. Those that meet the conditions are considered as outliers and removed. The judgment criteria are:

[0013] |x i -μ|>3σ;

[0014] Among them, x i is a single data point, μ is the mean of the data, and σ is the standard deviation of the data.

[0015] Furthermore, the method for performing data standardization processing on the operation data and the external environment data includes:

[0016] Standardize the cleaned data, adjust the variables of different ranges to the same scale, and improve the comparability of the data. The methods include:

[0017]

[0018] Among them, z i is the standardized data, x i is the original data, μ is the mean of the data, and σ is the standard deviation of the data.

[0019] Furthermore, the method for dynamically adjusting the digital twin of the photovoltaic power generation equipment using the external environment data, the operation data of the photovoltaic power generation equipment and the historical power generation data includes:

[0020] The output power of the digital twin of the photovoltaic power generation equipment is dynamically adjusted according to the light intensity. The output power P PV(s,t) =η pv ×S pv ×G (s,t) , where Ppv is the output power of photovoltaic power generation, η pv is the conversion efficiency of solar photovoltaic panels, S pv is the area of ​​photovoltaic panels (㎡), G (s,t) is the solar illuminance received at time t in season s (W / ㎡);

[0021] Dynamically adjust the photovoltaic conversion efficiency of the digital twin of photovoltaic power generation equipment according to temperature, in, is the reference temperature T ref Photovoltaic conversion efficiency under pv is the photovoltaic conversion efficiency at the actual working temperature; β is the temperature coefficient; T is the actual working temperature, T ref is the reference temperature;

[0022] The model parameters of the digital twin of photovoltaic power generation equipment are dynamically adjusted according to historical power generation data. The digital twin model dynamically adjusts the model parameters by real-time monitoring of light intensity and historical operation data of the equipment. Through multiple iterations, the digital twin gradually converges to a state that is highly consistent with the actual equipment behavior and provides support for future power generation forecasts.

[0023] Furthermore, the method for dynamically adjusting the digital twin of the wind power generation equipment using the external environment data, the operation data of the wind power generation equipment and the historical power generation data includes:

[0024] Dynamically adjust the output power of the digital twin of the wind power generation equipment according to wind speed and wind direction, and output power Where: P wind (t) represents the output power of the wind power generation equipment at time t; ρ is the air density, which is set to 1.225 kg / m 3 ; A is the area swept by the wind turbine blades; C p is the power coefficient, which indicates the efficiency of converting wind energy into mechanical energy; v(t) is the wind speed at time t;

[0025] Adjust the power coefficient of the wind turbine digital twin according to the blade design, wind speed and rotation speed of the wind turbine. Where: C pis the power coefficient, λ is the tip speed ratio, defined as the ratio of the blade tip speed to the wind speed; β is the pitch angle; e is the base of the natural logarithm;

[0026] The model parameters of the digital twin of wind turbines are dynamically adjusted according to historical power generation data. When there is a difference between the real-time monitoring data and the expected value of historical data, the twin model dynamically adjusts the model parameters based on feedback and uses an iterative optimization algorithm to gradually converge to a high degree of consistency with the actual equipment behavior, thereby improving the accuracy of future power generation forecasts.

[0027] Furthermore, the method of dynamically adjusting the digital twin of the energy storage device using the external environment data, the operation data of the energy storage device, and the historical operation data includes:

[0028] Adjust the power output of the digital twin of the energy storage device according to the operating data of the energy storage device. Where: P storage (t) is the output power of the energy storage device at time t, η discharge is the discharge efficiency; η charge is the charging efficiency; P out (t) is the discharge power of the device at time t; P in (t) is the charging power of the device at time t;

[0029] Dynamically calculate the remaining storage of the digital twin of the energy storage device, the remaining storage Where: E storage (t) is the remaining capacity of the energy storage device at time t; E storage (t-1) is the remaining capacity at the previous moment; Δt is the time interval;

[0030] The digital twin of the energy storage device is dynamically adjusted based on historical operating data. By combining the historical operating data and real-time data of the device, real-time monitoring of key parameters such as charging and discharging status, storage capacity, and device temperature is performed to reflect its operating status. When there is a deviation between the monitoring data and the prediction of the historical model, the twin model dynamically adjusts parameters based on feedback.

[0031] A device-level digital twin construction system based on model data fusion, used to execute a device-level digital twin construction method based on model data fusion, characterized in that it includes:

[0032] Distributed sensor network for real-time collection of operating data of photovoltaic power generation, wind power generation and energy storage equipment: voltage, current and power output; also for collecting external environmental data: temperature, light intensity and wind speed;

[0033] A central control system, used to receive data acquired by a distributed sensor network, and to perform data cleaning and data standardization processing on the operation data and the external environment data;

[0034] A photovoltaic digital twin system, for receiving data processed by a central control system, for constructing a digital twin of a photovoltaic power generation device, and for dynamically adjusting the digital twin of the photovoltaic power generation device using the external environment data, the operation data of the photovoltaic power generation device, and the historical power generation data;

[0035] A wind power digital twin system, for receiving data processed by a central control system, for building a digital twin of a wind power generation equipment, and for dynamically adjusting the digital twin of the wind power generation equipment using the external environment data, the operation data of the wind power generation equipment, and the historical power generation data;

[0036] The energy storage digital twin system is used to receive data processed by the central control system, to build a digital twin of the energy storage device, and to dynamically adjust the digital twin of the energy storage device using the external environment data, the operating data of the energy storage device, and the historical operating data.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1) Through multi-source data fusion, refined modeling and dynamic optimization of equipment can be achieved to ensure accurate reflection and regulation of equipment status.

[0039] 2) Through multi-source data fusion, refined modeling and dynamic optimization of equipment can be achieved to ensure accurate reflection and regulation of equipment status.

[0040] 3) Utilize the combination of real-time data and historical data to achieve accurate optimization of equipment operation strategies and improve the robustness and response speed of the system BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0042] Figure 1 : Typical photovoltaic output curve diagram in Example 1 of the present invention;

[0043] Figure 2 : Typical output curve of the fan in Example 1 of the present invention; DETAILED DESCRIPTION

[0044] In order to better understand the present invention, the content of the present invention is further clearly described below in conjunction with the embodiments and the accompanying drawings, but the protection content of the present invention is not limited to the following embodiments. In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details.

[0045] Embodiment 1: This embodiment is a device-level digital twin construction method based on model data fusion. The present invention takes device-level digital twin construction as the core, and improves the modeling accuracy and adaptive optimization capability of distributed energy equipment through multivariate data fusion and real-time interaction technology. The present invention proposes a digital twin construction method for photovoltaic power generation, wind power generation and energy storage equipment, and combines real-time monitoring data, historical data and virtual simulation to construct a device-level twin model with high precision and dynamic adaptability.

[0046] This embodiment includes:

[0047] S1. Collection and preprocessing of equipment-level operation data

[0048] The present invention first collects the operating data of photovoltaic power generation, wind power generation and energy storage equipment in real time through a distributed sensor network. The collected data includes electrical quantity data such as voltage, current, power output, and external environmental data such as temperature, light intensity, wind speed, etc. The system transmits various types of data to the central control system through the Internet of Things technology. In order to ensure the reliability and consistency of the data, the present invention adopts data cleaning and standardization processing technology. Specifically, it includes the following steps:

[0049] Step 1: Data cleaning. In the data cleaning process, firstly, outliers in the data are removed by using an outlier detection algorithm. Outliers can be detected by the 3σ principle, that is, the mean and standard deviation of the collected data are analyzed, and those that meet the conditions are regarded as outliers and removed. The formula is as follows:

[0050] |x i -μ|>3σ (1)

[0051] Where xi is a single data point, μ is the mean of the data, and σ is the standard deviation of the data. If a data point exceeds 3 times the standard deviation of the mean, it is considered abnormal data.

[0052] Step 2: Data standardization. After data cleaning, the system standardizes the data. Standardization is processed through the Z-score standardization formula to ensure that different types of data are represented in the same range to facilitate subsequent modeling and analysis:

[0053]

[0054] Among them, zi is the standardized data, xi is the original data, μ is the mean of the data, and σ is the standard deviation of the data. This standardization method can adjust variables of different ranges to the same scale, effectively improving the comparability of the data.

[0055] After cleaning and standardization, the system obtains standardized equipment operation data that eliminates noise and outliers. These data provide an accurate and consistent basis for subsequent digital twin modeling.

[0056] S2. Digital twin modeling of photovoltaic power generation equipment

[0057] like Figure 1 As shown, the digital twin modeling of photovoltaic power generation equipment is based on the physical characteristics and environmental factors of the equipment to construct a virtual model that can reflect the dynamic behavior of photovoltaic power generation in real time. The present invention first obtains the electrical quantity data (such as voltage, current, output power) and external environmental data (such as light intensity and temperature) of photovoltaic power generation equipment in real time through distributed sensors. These data are transmitted to the digital twin system through the Internet of Things for real-time monitoring of equipment status and model update.

[0058] (1) Output power model

[0059] The output power Ppv of photovoltaic power generation is closely related to factors such as the conversion efficiency of photovoltaic panels, light intensity and the area of ​​photovoltaic panels. The basic output power calculation formula is:

[0060] P PV(s,t) =η pv ×S pv ×G (s,t) (3)

[0061] Where: η pv is the conversion efficiency of solar photovoltaic panels, S pv is the area of ​​photovoltaic panels (㎡), G (s,t) is the solar illuminance received at time t in season s (W / ㎡).

[0062] The digital twin model dynamically adjusts model parameters by real-time monitoring of light intensity G(t) and historical equipment operation data. Specifically, the power generation capacity of photovoltaic power generation equipment is greatly affected by external climate change, so the model needs to be dynamically optimized according to real-time changes in environmental parameters such as light intensity and temperature. Through multiple iterations, the digital twin can gradually converge to a state that is highly consistent with the actual equipment behavior and provide accurate support for future power generation forecasts.

[0063] (2) Temperature effect correction

[0064] The output of photovoltaic power generation equipment is not only affected by light intensity, but temperature is also an important factor affecting photovoltaic conversion efficiency. The effect of temperature on photovoltaic cell efficiency can be characterized by the following empirical formula:

[0065]

[0066] in, is the reference temperature T ref β is the temperature coefficient (usually a negative value, indicating that the efficiency decreases when the temperature increases); T is the actual operating temperature, T ref is the reference temperature (usually 25°C). This formula is used to adjust the power output of the twin model at different ambient temperatures.

[0067] The above formula is applied to correct the output efficiency of photovoltaic equipment, thereby improving the accuracy of the model.

[0068] (3) Historical data and dynamic adjustment

[0069] The digital twin model combines the historical power generation data of the equipment with real-time sensor data to achieve dynamic adjustment. Historical data modeling first collects a large amount of historical data, including light intensity G(s,t), temperature T, and output power P pv , and model these data through regression analysis, machine learning or deep learning algorithms to extract the influence of environmental conditions on equipment performance. In the process of real-time adjustment and iterative optimization, when there is a difference between the expected value of real-time monitoring data and historical data, the twin model dynamically adjusts the model parameters based on feedback, and gradually converges to a high degree of consistency with the actual equipment behavior using an iterative optimization algorithm, thereby improving the accuracy of future power generation forecasts. Through real-time data input of light and temperature, the twin model can predict the output power at subsequent moments and adjust the output power P according to real-time temperature and light changes. pv , for example, to instantly correct η when light intensity decreases or temperature rises pv , to ensure the accuracy of the prediction results.

[0070] S3. Digital twin modeling of wind power equipment

[0071] like Figure 2 As shown, the digital twin modeling of wind power generation equipment is based on the physical characteristics and environmental factors of the equipment to build a virtual model that can reflect the dynamic behavior of wind power generation in real time. The present invention obtains key operating data (such as wind speed, wind direction, blade speed) and external environmental data of wind power generation equipment in real time through a distributed sensor network. These data are transmitted to the digital twin system through the Internet of Things for real-time monitoring of equipment status and model updating.

[0072] (1) Output power model

[0073] The output power of wind power generation P wind It is closely related to factors such as wind speed, air density, blade swept area and power coefficient. The basic output power calculation formula is:

[0074]

[0075] Where: P wind (t) represents the output power of the wind power generation equipment at time t; ρ is the air density, which is usually 1.225 kg / m 3 ; A is the area swept by the wind turbine blades; C p is the power coefficient, which indicates the efficiency of converting wind energy into mechanical energy; v(t) is the wind speed at time t.

[0076] When the wind speed is lower than the cut-in wind speed or higher than the cut-out wind speed, the wind turbine will stop working, so the power output is zero. The output of wind power generation is greatly affected by the fluctuation of wind speed. The twin model monitors the wind speed changes in real time and dynamically adjusts the operating parameters of the power generation equipment to ensure that the system operates efficiently under different wind speed conditions.

[0077] (2) Power factor adjustment

[0078] The efficiency of wind power generation is also affected by the power coefficient C p The power coefficient has a complex nonlinear relationship with the blade design, wind speed and rotation speed of the wind turbine. Usually, the power coefficient can be represented by the following empirical formula:

[0079]

[0080] Where: λ is the tip speed ratio, defined as the ratio of the blade tip speed to the wind speed; β is the pitch angle; and e is the base of the natural logarithm.

[0081] By monitoring the blade speed and wind speed in real time, the twin model can dynamically calculate the power coefficient C p , and adjusts the tip speed ratio λ according to the real-time wind speed changes and rotation speed to optimize the power generation efficiency. This formula is used to adjust the power output under different wind speed conditions to ensure efficient operation of the equipment.

[0082] (3) Historical data and dynamic adjustment

[0083] The digital twin model of wind power equipment combines the historical operation data of the equipment with real-time sensor data to achieve dynamic adjustment. Historical data modeling first collects a large amount of historical data, including wind speed v(t), wind direction, blade speed and output power P wind, and model these data through regression analysis, machine learning or deep learning algorithms to extract the influence of environmental conditions on equipment performance. In the process of real-time adjustment and iterative optimization, when there is a difference between the expected value of real-time monitoring data and historical data, the twin model dynamically adjusts the model parameters based on feedback, and gradually converges to a high degree of consistency with the actual equipment behavior using an iterative optimization algorithm, thereby improving the accuracy of future power generation forecasts. Through real-time data input of wind speed and blade speed, the twin model can predict the output power at subsequent times and adjust the power output P according to the real-time wind speed changes. wind , for example, to correct C immediately when the wind speed suddenly increases or decreases p , to ensure the accuracy of the prediction results.

[0084] S4. Digital twin modeling of energy storage equipment

[0085] Energy storage equipment plays a key role in distributed energy systems, responsible for releasing electricity when electricity demand peaks and storing electricity when there is excess electricity. The digital twin modeling of energy storage equipment can accurately reflect its operating status through real-time monitoring of key parameters such as charging and discharging status, storage capacity, and equipment temperature, and dynamically adjust the charging and discharging strategy of energy storage equipment according to system requirements.

[0086] (1) Power balance during charging and discharging

[0087] The charging and discharging process of energy storage equipment can be expressed by the power balance equation. The power output of the energy storage device at time t is affected by the current state of the battery, load demand and system operation strategy. Its basic power output formula is:

[0088]

[0089] Where: η discharge is the discharge efficiency; η charge is the charging efficiency; P out (t) is the discharge power of the device at time t; P in (t) is the charging power of the device at time t.

[0090] (2) Dynamic capacity calculation

[0091] The remaining storage capacity of the energy storage device is dynamically calculated through the charging and discharging process of the device over a period of time. The capacity change of the energy storage device follows the following formula:

[0092]

[0093] Where: E storage (t) is the remaining capacity of the energy storage device at time t; E storage(t-1) is the remaining capacity at the previous moment; Δt is the time interval. This formula indicates that the storage capacity of the energy storage device at time t is the capacity at the previous time point plus the charge amount during this period minus the discharge amount. Through this model, the energy storage device can dynamically adjust the charging and discharging strategy according to the current system load demand and the actual operation of the equipment.

[0094] (3) Historical data and dynamic adjustment

[0095] The digital twin model of energy storage equipment combines the historical operating data of the equipment with real-time sensor data to achieve dynamic adjustment and optimization. First, the system collects historical data such as charge and discharge status, storage capacity, temperature, etc., and extracts the influence of environmental conditions on energy storage performance through regression analysis, machine learning or deep learning algorithms. During the real-time adjustment process, when there is a deviation between the monitoring data and the prediction of the historical model, the twin model dynamically adjusts the parameters based on feedback to ensure equipment operation efficiency and system balance.

[0096] (4) Load demand and strategy optimization

[0097] During operation, the twin model continuously monitors the power load and the status of energy storage devices, and adjusts the charging and discharging strategy in real time to ensure the balance of system supply and demand and maximize the utilization of energy storage devices. For example, when the system load demand increases, the twin model gives priority to using energy storage devices for discharge to reduce system pressure; when the load demand decreases, the model decides whether to charge based on the electricity price or clean energy supply. In addition, the model can analyze future load demand trends based on historical data, optimize charging and discharging in advance, and improve the operating efficiency of the equipment.

[0098] Beneficial Effects

[0099] 1: Equipment-level operation data collection and preprocessing, to obtain real-time operation data of photovoltaic power generation, wind power generation and energy storage equipment. The data includes electrical quantity data such as voltage, current, power output, and external environmental data such as light intensity, wind speed, and temperature. The collected data is cleaned and standardized to eliminate outliers and ensure that the data is processed on the same scale to ensure the accuracy of subsequent modeling;

[0100] 2: For photovoltaic power generation equipment, combine its characteristics with environmental data to build a digital twin model of photovoltaic power generation equipment. By real-time monitoring of data such as light intensity and temperature of the equipment, the model can dynamically reflect the power generation capacity of the photovoltaic equipment and adjust the equipment's operating strategy based on real-time data to ensure efficient operation of the equipment;

[0101] 3: For wind power generation equipment, a digital twin model of the wind power generation equipment is constructed by combining its wind speed, wind direction, blade speed and other data. The model adjusts the operating parameters of the wind power generation equipment according to the real-time wind speed and speed to optimize the power generation efficiency;

[0102] 4: For energy storage equipment, a digital twin model of the energy storage equipment is built to monitor the charging and discharging status, storage capacity and temperature data of the equipment. The model dynamically adjusts the charging and discharging strategy according to the system load demand and equipment status to ensure the efficient use of the equipment in the system.

[0103] Embodiment 2: A device-level digital twin construction system based on model data fusion, used to execute the device-level digital twin construction method based on model data fusion described in Embodiment 1, comprising:

[0104] Distributed sensor network for real-time collection of operating data of photovoltaic power generation, wind power generation and energy storage equipment. Electrical quantity data: voltage, current and power output; also for collecting external environment data: temperature, light intensity and wind speed;

[0105] The central control system is used to receive the operating data acquired by the distributed sensor network, and to perform data cleaning and data standardization on the operating data and the external environment data; in combination with Example 1, the distributed sensor network and the central control system are used to execute step S1.

[0106] A photovoltaic digital twin system is used to receive data collected by a distributed sensor network, integrate the data with the physical characteristics of wind turbines or photovoltaic power generation, and construct a digital twin that reflects the operating status of the corresponding equipment. The data processed by the central control system is used to monitor and predict the equipment status in real time and update the model based on the data to construct a digital twin of the photovoltaic power generation equipment, and to dynamically adjust the digital twin of the photovoltaic power generation equipment using the external environment data, the operating data of the photovoltaic power generation equipment and the historical power generation data; in combination with Example 1, the photovoltaic digital twin system is used to execute step S2.

[0107] A wind power digital twin system is used to receive data processed by a central control system, to construct a digital twin of a wind power generation equipment, and to dynamically adjust the digital twin of the wind power generation equipment using the external environment data, the operating data of the wind power generation equipment, and the historical power generation data; in combination with Example 1, the wind power digital twin system is used to execute step S3.

[0108] The energy storage digital twin system is used to receive data processed by the central control system, to construct a digital twin of the energy storage device, and to dynamically adjust the digital twin of the energy storage device using the external environment data, the operating data of the energy storage device, and the historical operating data; in combination with Example 1, the energy storage digital twin system is used to execute step S4.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Other modifications or equivalent substitutions made to the technical solution of the present invention by ordinary technicians in the field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solution of the present invention.

Claims

1. A device-level digital twin construction method based on model data fusion, characterized in that: include: Collect operating data and external environment data of photovoltaic power generation, wind power generation and energy storage equipment; Performing data cleaning and data standardization processing on the operation data and the external environment data; Constructing a digital twin of the photovoltaic power generation equipment, and dynamically adjusting the digital twin of the photovoltaic power generation equipment using the external environment data, the operation data of the photovoltaic power generation equipment, and the historical power generation data; Constructing a digital twin of the wind power generation equipment, and dynamically adjusting the digital twin of the wind power generation equipment using the external environment data, the operation data of the wind power generation equipment, and the historical power generation data; A digital twin of the energy storage device is constructed, and the digital twin of the energy storage device is dynamically adjusted using the external environment data, the operating data of the energy storage device, and the historical operating data.

2. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: Methods for collecting operation data and external environment data of photovoltaic power generation, wind power generation and energy storage equipment include: Real-time collection of operating data of photovoltaic power generation, wind power generation and energy storage equipment. The collected electrical quantity data include voltage, current and power output; the collected external environment data include temperature, light intensity and wind speed.

3. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: The method for performing data cleaning on the operation data and the external environment data includes: The outliers in the data are removed by using the outlier detection algorithm. The outliers are detected by the 3σ principle. The mean and standard deviation of the collected data are analyzed. Those that meet the conditions are considered as outliers and removed. The judgment criteria are: |x i -μ|>3σ; Among them, x i is a single data point, μ is the mean of the data, and σ is the standard deviation of the data.

4. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: The method for performing data standardization processing on the operation data and the external environment data includes: Standardize the cleaned data, adjust the variables of different ranges to the same scale, and improve the comparability of the data. The methods include: Among them, z i is the standardized data, x i is the original data, μ is the mean of the data, and σ is the standard deviation of the data.

5. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: The method for dynamically adjusting the digital twin of the photovoltaic power generation equipment by using the external environment data, the operation data of the photovoltaic power generation equipment and the historical power generation data includes: The output power of the digital twin of the photovoltaic power generation equipment is dynamically adjusted according to the light intensity. The output power P PV(s,t) =η pv ×S pv ×G (s,t) , where Ppv is the output power of photovoltaic power generation, η pv is the conversion efficiency of solar photovoltaic panels, S pv is the area of ​​photovoltaic panels (㎡), G (s,t) is the solar illuminance received at time t in season s (W / ㎡); Dynamically adjust the photovoltaic conversion efficiency of the digital twin of photovoltaic power generation equipment according to temperature, in, is the reference temperature T ref Photovoltaic conversion efficiency under pv is the photovoltaic conversion efficiency at the actual working temperature; β is the temperature coefficient; T is the actual working temperature, T ref is the reference temperature; The model parameters of the digital twin of photovoltaic power generation equipment are dynamically adjusted according to historical power generation data. The digital twin model dynamically adjusts the model parameters by real-time monitoring of light intensity and historical operation data of the equipment. Through multiple iterations, the digital twin gradually converges to a state that is highly consistent with the actual equipment behavior and provides support for future power generation forecasts.

6. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: The method for dynamically adjusting the digital twin of the wind power generation equipment by using the external environment data, the operation data of the wind power generation equipment and the historical power generation data includes: Dynamically adjust the output power of the digital twin of the wind power generation equipment according to wind speed and wind direction, and output power Where: P wind (t) represents the output power of the wind power generation equipment at time t; ρ is the air density, which is set to 1.225 kg / m 3 ; A is the area swept by the wind turbine blades; C p is the power coefficient, which indicates the efficiency of converting wind energy into mechanical energy; v(t) is the wind speed at time t; Adjust the power coefficient of the wind turbine digital twin according to the blade design, wind speed and rotation speed of the wind turbine. Where: C p is the power coefficient, λ is the tip speed ratio, defined as the ratio of the blade tip speed to the wind speed; β is the pitch angle; e is the base of the natural logarithm; The model parameters of the digital twin of wind turbines are dynamically adjusted according to historical power generation data. When there is a difference between the real-time monitoring data and the expected value of historical data, the twin model dynamically adjusts the model parameters based on feedback and uses an iterative optimization algorithm to gradually converge to a high degree of consistency with the actual equipment behavior, thereby improving the accuracy of future power generation forecasts.

7. The device-level digital twin construction method based on model data fusion according to claim 1, characterized in that: The method for dynamically adjusting the digital twin of the energy storage device using the external environment data, the operation data of the energy storage device, and the historical operation data includes: Adjust the power output of the digital twin of the energy storage device according to the operating data of the energy storage device. Where: P storage (t) is the output power of the energy storage device at time t, η discharge is the discharge efficiency; η charge is the charging efficiency; P out (t) is the discharge power of the device at time t; P in (t) is the charging power of the device at time t; Dynamically calculate the remaining storage of the digital twin of the energy storage device, the remaining storage Where: E storage (t) is the remaining capacity of the energy storage device at time t; E storage (t-1) is the remaining capacity at the previous moment; Δt is the time interval; The digital twin of the energy storage device is dynamically adjusted based on historical operating data. By combining the historical operating data and real-time data of the device, real-time monitoring of key parameters such as charging and discharging status, storage capacity, and device temperature is performed to reflect its operating status. When there is a deviation between the monitoring data and the prediction of the historical model, the twin model dynamically adjusts parameters based on feedback.

8. A device-level digital twin construction system based on model data fusion, used to execute the device-level digital twin construction method based on model data fusion according to any one of claims 1 to 7, characterized in that: include: Distributed sensor networks for real-time collection of operating data from photovoltaic, wind power and energy storage equipment: voltage, current and power output; It is also used to collect external environmental data: air temperature, light intensity and wind speed; A central control system, used to receive data acquired by a distributed sensor network, and to perform data cleaning and data standardization processing on the operation data and the external environment data; A photovoltaic digital twin system, for receiving data processed by a central control system, for constructing a digital twin of a photovoltaic power generation device, and for dynamically adjusting the digital twin of the photovoltaic power generation device using the external environment data, the operation data of the photovoltaic power generation device, and the historical power generation data; A wind power digital twin system, for receiving data processed by a central control system, for building a digital twin of a wind power generation equipment, and for dynamically adjusting the digital twin of the wind power generation equipment using the external environment data, the operation data of the wind power generation equipment, and the historical power generation data; The energy storage digital twin system is used to receive data processed by the central control system, to build a digital twin of the energy storage device, and to dynamically adjust the digital twin of the energy storage device using the external environment data, the operating data of the energy storage device, and the historical operating data.

Images