Multi-energy station reliability simulation method and system based on digital twinning

Abstract

The present application relates to the technical field of simulation and analog, and proposes a multi-energy station reliability simulation method and system based on digital twinning, which comprises the following steps: collecting the energy flow power of each energy station at each collection time and the environmental parameters of the multi-energy station at each collection time, establishing the environmental parameter sequence and the energy flow power sequence; calculating the multi-energy flow disturbance level and the reliability evaluation value at the collection time; and then determining the convergence tolerance at the collection time, and carrying out multi-energy station reliability simulation analysis through the convergence tolerance. The present application can improve the reliability of the simulation and analog results.

Description

Technical Field

[0001] This invention relates to the field of simulation technology, specifically to a method and system for simulating the reliability of multi-energy power plants based on digital twins. Background Technology

[0002] Generally, digital twin technology is used to build a high-fidelity real-time mapping system between physical power stations and virtual models. This system relies on virtual iterative testing and integrates IoT-based data collection with mechanistic simulation analysis to simulate the power station's operational status. However, in actual operation, the power output of wind and solar power at the power station is easily affected by instantaneous fluctuations in weather conditions, and the aging and wear of various operating equipment exhibit nonlinear, time-varying drift characteristics. Various uncertainties are transmitted, superimposed, and amplified layer by layer through multi-energy flow coupling links (electricity, heat, and gas), ultimately forming a high-dimensional and complex uncertainty propagation network.

[0003] Currently, the Monte Carlo method is commonly used for power plant reliability analysis. However, due to the strong nonlinear characteristics of multi-energy flow coupling and the propagation of high-dimensional uncertainties, existing surrogate models struggle to accurately characterize the complex dynamic operation of power plants. Specifically, conventional surrogate models such as multinomial chaos and Bayesian models all have inherent prior constraint assumptions. Furthermore, when faced with complex multi-source input data, their fitting accuracy is low, directly leading to a large confidence interval range for reliability assessment indicators. This significantly reduces the reference value of the assessment results, making it difficult to support the formulation of scientific and reasonable equipment scheduling schemes and the implementation of risk control measures. This severely restricts the improvement of the safe, stable, efficient, and economical operation and maintenance level of multi-energy power plants. Summary of the Invention

[0004] This invention provides a method and system for simulating the reliability of multi-energy power stations based on digital twins, in order to solve the problem that the simulation results of multi-energy power stations are unreliable due to the influence of meteorological conditions and aging losses. The specific technical solution adopted is as follows: In a first aspect, one embodiment of the present invention provides a multi-energy power station reliability simulation method based on digital twins, the method comprising the following steps: Collect the energy flow power of each energy station at each collection time, and the environmental parameters of multiple energy stations at each collection time, and establish the environmental parameter sequence and the energy flow power sequence of the energy station at the collection time. The energy flow includes electrical energy flow, thermal energy flow and natural gas energy flow. The environmental parameters include ambient wind speed, irradiance and temperature. Based on the empirical mode decomposition results of the environmental parameter sequence at the time of acquisition, and the coupling relationship between the power sequences of different energy flows at the time of acquisition, the multi-energy flow disturbance level at the time of acquisition is calculated, and the multi-energy flow disturbance level sequence at the time of acquisition is established. Based on the persistence and extreme value distribution characteristics of the multi-energy flow disturbance levels within the multi-energy flow disturbance level sequence at the time of acquisition, the reliability assessment value at the time of acquisition is calculated. Based on the reliability assessment value at the time of data acquisition, the convergence tolerance at the time of data acquisition is determined, and the reliability simulation analysis of multi-energy power stations is carried out through the convergence tolerance.

[0005] Furthermore, the data in both the environmental parameter sequence and the energy flow power sequence are normalized.

[0006] Furthermore, the method for calculating the multi-energy flow disturbance level at the acquisition time is as follows: Based on the empirical mode decomposition results of the environmental parameter sequence at the time of acquisition, the proportion of high-frequency energy in the environmental parameter sequence is calculated. The coupling dependency difference at different time points is calculated based on the coupling relationship between the power sequences of different energy flows at the time of acquisition. The product of the arithmetic mean of the high-frequency energy proportions of all environmental parameter sequences at the acquisition time and the coupling dependency difference is denoted as the multi-energy flow disturbance level at the acquisition time.

[0007] Furthermore, the high-frequency energy proportion of the environmental parameter sequence is specifically as follows: Empirical mode decomposition (EMD) is performed on the environmental parameter sequence at the acquisition time to obtain all IMF components. The ratio of the energy of the two highest-frequency IMF components to the total energy of all IMF components is recorded as the high-frequency energy ratio of the environmental parameter sequence.

[0008] Furthermore, the specific calculation method for the coupling dependency difference at the acquisition time is as follows: For the same acquisition time, the mutual information value of the power sequences of the two different energy streams is calculated, and the coefficient of variation of the mutual information value of the power sequences of the two different energy streams is recorded as the coupling dependency difference at the corresponding acquisition time.

[0009] Furthermore, the method for calculating the reliability assessment value at the acquisition time is as follows: Calculate the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time; Based on the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time and the values ​​of all multi-energy flow disturbance levels within the multi-energy flow disturbance level sequence, the POT threshold is set, and the shape parameter and over-threshold frequency at the acquisition time are calculated based on the multi-energy flow disturbance level sequence at the acquisition time. The reliability assessment value at the acquisition time is calculated based on the over-threshold frequency and shape parameters at the acquisition time.

[0010] Furthermore, the method for setting the POT threshold is as follows: If the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time is less than or equal to the preset persistence threshold, the POT threshold is the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 1.5 times the standard deviation; otherwise, the POT threshold is the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 2 times the standard deviation.

[0011] Furthermore, the specific method for calculating the reliability assessment value at the acquisition time based on the over-threshold frequency and shape parameters at the acquisition time includes: The frequency exceeding the threshold at the acquisition time is used as the numerator, and the sum of the mean of the frequencies exceeding the threshold at all acquisition times before the acquisition time and the preset adjustment coefficient is used as the denominator. The value of the fraction is recorded as the relative frequency exceeding the threshold at the acquisition time. The maximum value between the shape parameter and the number 0 at the acquisition time is selected, and the product of the selected maximum value and the relative frequency exceeding the threshold at the acquisition time is recorded as the risk characteristic value at the acquisition time. The negative of the risk characteristic value at the acquisition time is used as the exponent value, and the value of the exponential function with the natural constant as the base is recorded as the reliability assessment value at the acquisition time.

[0012] Furthermore, the method for calculating the convergence tolerance at the acquisition time is as follows: The difference between the number 1 and the reliability assessment value at the acquisition time is used as the numerator. The sum of the product of the preset saturation coefficient and the reliability assessment value at the acquisition time and the number 1 is used as the denominator. The product of the value of the fraction and the preset adjustment coefficient is recorded as the characteristic product at the acquisition time. The product of the difference between the number 1 and the characteristic product at the acquisition time and the preset initial tolerance is recorded as the convergence tolerance at the acquisition time.

[0013] Secondly, embodiments of the present invention also provide a multi-energy power station reliability simulation system based on digital twins, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.

[0014] The beneficial effects of this invention are: This application first evaluates the possibility of the energy flow being in a severely non-stationary state based on the relative intensity of rapid random disturbances caused by environmental changes in the energy flow of a multi-energy power station and the degree of difference in the dependence strength between different coupling channels. It analyzes the two major factors leading to system instability—the severe random disturbances of the energy flow itself and the non-equilibrium distribution of coupling strength between energy flows—to obtain the coupling dependence differences at each data acquisition time and establish a multi-energy flow disturbance level sequence at each acquisition time. Furthermore, considering the long-term memory and nonlinear transmission characteristics of disturbance coupling between different energy flows in a multi-energy power station, it analyzes the possibility of the multi-energy power station being in a reliable state and obtains the reliability assessment value at each acquisition time. Finally, based on the reliability assessment value at each acquisition time, it determines the convergence tolerance at each acquisition time and conducts reliability simulation analysis of the multi-energy power station through the convergence tolerance, addressing the problem of unreliable simulation results caused by the influence of meteorological conditions and aging losses on multi-energy power stations. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a multi-energy power station reliability simulation method based on digital twins, provided in one embodiment of the present invention. Detailed Implementation

[0016] Please see Figure 1 The diagram illustrates a flowchart of a multi-energy power station reliability simulation method based on digital twins, according to an embodiment of the present invention. The method includes the following steps: Step S001: Collect the energy flow power of each energy station at each collection time and the environmental parameters of multiple energy stations at each collection time, and establish the environmental parameter sequence and the energy flow power sequence of the energy station at the collection time. The energy flow includes electrical energy flow, thermal energy flow and natural gas energy flow. The environmental parameters include ambient wind speed, irradiance and temperature.

[0017] Among them, energy stations include, but are not limited to, wind power stations, photovoltaic power stations, energy storage stations, gas turbine stations and heating stations, and energy flows include, but are not limited to, electrical energy flows, heat energy flows and natural gas energy flows.

[0018] For electrical energy flows, current transformers and voltage transformers are deployed at the grid connection point of the corresponding energy station to collect the instantaneous current and voltage of the electrical energy flow. Combined with the power factor, the active power of the electrical energy flow is calculated, i.e., the energy flow power of the electrical energy flow. For thermal energy flows, heat meters are installed at the outlet of the corresponding energy station to measure the mass flow rate of the heat transfer fluid and the supply and return water temperature difference, calculating the thermal power of the thermal energy flow, i.e., the energy flow power of the thermal energy flow. For natural gas energy flows, gas flow meters are installed at the gas pipeline connection point of the gas turbine station of the corresponding energy station to collect the flow rate. The average calorific value of the gas is continuously detected and calculated using a calorific value meter. The product of the volumetric flow rate and the calorific value of the gas is recorded as the gas power of the natural gas energy flow, i.e., the energy flow power of the natural gas energy flow.

[0019] An anemometer was installed at the wind turbine hub, an irradiance meter was installed at the tilt angle of the photovoltaic module array plane, and a temperature sensor was installed at an unobstructed location in the center of the power station to collect ambient wind speed, irradiance, and temperature data in the multi-energy power station. Ambient wind speed, irradiance, and temperature were all recorded as environmental parameters.

[0020] In this embodiment, the sampling frequency for energy flow power, ambient wind speed, irradiance, and temperature is all set to 10Hz.

[0021] For each data acquisition moment, the normalized values ​​of the power of all energy flows of the same type or the normalized values ​​of environmental parameters within the preset time window before the acquisition moment are arranged sequentially to establish an energy flow power sequence or an environmental parameter sequence.

[0022] In this embodiment, the duration of the preset time window is set to 1 hour. This embodiment uses the maximum-minimum normalization method to calculate the normalized value. In the maximum-minimum normalization method, the sum of the difference between the maximum and minimum values ​​and a preset non-zero small constant is used as the denominator to calculate the normalized value. In this embodiment, the preset non-zero small constant is set to a value of... In practical applications, implementers may use other methods from existing technologies, such as the tanh function or the sigmoid function, to calculate the normalized value; no restrictions are imposed here.

[0023] Thus, the energy flow power sequence of each energy station at each acquisition time, as well as the environmental parameter sequence at each acquisition time, are obtained.

[0024] Step S002: Based on the empirical mode decomposition results of the environmental parameter sequence at the acquisition time and the coupling relationship between the energy flow power sequences of different energy flows at the acquisition time, calculate the multi-energy flow disturbance level at the acquisition time and establish the multi-energy flow disturbance level sequence at the acquisition time.

[0025] In multi-energy power plants, wind and solar power output is easily affected by instantaneous weather fluctuations. Traditional analysis methods cannot simultaneously depict the coupling relationship between various energy flows. When relying on digital twin models to carry out reliability simulations, it is difficult to accurately identify the source of disturbance and the path of uncertainty propagation.

[0026] Empirical mode decomposition (EMD) is performed on the environmental parameter sequence at the acquisition time to obtain all IMF components. The energy of the IMF components is calculated, and the top two IMF components with the highest frequency are taken as high-frequency components. The ratio of the energy of all high-frequency components to the total energy of all IMF components is recorded as the high-frequency energy ratio of the environmental parameter sequence.

[0027] Among them, empirical mode decomposition and the calculation of the energy of IMF components are well-known techniques and will not be described in detail. In this embodiment, the environmental parameter sequence is decomposed into 8th-order IMF components, and the energy of the IMF component is the sum of squares of all data in the IMF component.

[0028] Empirical Mode Decomposition (EMD) is used to adaptively decompose nonlinear, non-stationary signals into a predetermined number of Intrinsic Mode Fractions (IMFs). Each IMF represents a rapidly fluctuating component in the input signal, such as random noise, and its energy proportion quantifies the relative intensity of the input signal being subjected to rapid random disturbances. The high-frequency energy proportion of the environmental parameter sequence indicates the relative intensity of rapid random disturbances caused by environmental changes to the energy flow of multiple energy stations at the time of acquisition and within a predetermined time window prior to the acquisition time. A higher high-frequency energy proportion in the environmental parameter sequence indicates that environmental changes are more likely to cause the energy flow to be in a drastic non-stationary state, and the digital twin model needs to focus more on dynamic evolution.

[0029] The operational reliability of multi-energy power plants depends on the synergistic and complementary characteristics of electrical, thermal, and gas energy flows and the fault propagation patterns. Simply analyzing the high-frequency disturbances of the energy flow itself cannot effectively reveal the disturbance propagation patterns and coupling effects between energy flows.

[0030] For the same acquisition time, the mutual information value of the power sequences of the two different energy streams is calculated, and the coefficient of variation of the mutual information value of the power sequences of the two different energy streams is recorded as the coupling dependency difference at the corresponding acquisition time.

[0031] The calculation of mutual information value is a well-known technique and will not be elaborated further.

[0032] Mutual information is used to analyze the correlation between two variables. The larger the mutual information value, the stronger the coupling relationship between the two variables. When the difference in coupling dependency at the time of acquisition is greater, the difference in dependency strength between different coupling channels is also greater, and local disturbances are more likely to cause global instability through a few strong coupling chains.

[0033] The product of the arithmetic mean of the high-frequency energy proportions of all environmental parameter sequences at the acquisition time and the coupling dependency difference is denoted as the multi-energy flow disturbance level at the acquisition time.

[0034] In multi-energy power plants, the severe random disturbances of the energy flow itself and the non-equilibrium distribution of coupling strength between energy flows are the two major factors leading to system instability. The greater the disturbance level of the multi-energy flow at the time of acquisition, the more likely the energy flow is to be in a severe non-stationary state at the time of acquisition, and the more likely global instability will occur at the time of acquisition.

[0035] It should be noted that the high-frequency energy proportion and coupling dependency difference are not completely independent in actual engineering scenarios. When a single-source disturbance occurs in the system, the disturbance will be transmitted through the multi-energy flow coupling link, causing changes in both the high-frequency energy proportion and the coupling dependency difference simultaneously. There will not be a situation where one indicator significantly increases while the other strictly approaches zero. This application uses the product of the two to characterize the level of multi-energy flow disturbance, in order to depict the global disturbance propagation risk caused by the combined effects of rapid environmental disturbances and coupling link imbalances.

[0036] All multi-energy flow disturbance levels within the preset time window before the acquisition time are arranged sequentially to establish a multi-energy flow disturbance level sequence at the acquisition time.

[0037] Thus, the multi-energy flow disturbance level sequence at the acquisition time is obtained.

[0038] Step S003: Calculate the reliability assessment value at the time of acquisition based on the persistence and extreme value distribution characteristics of the multi-energy flow disturbance level within the multi-energy flow disturbance level sequence at the acquisition time.

[0039] In multi-energy power plants, the disturbance coupling between different energy flows has long-term memory and nonlinear transmission characteristics. Traditional methods have difficulty distinguishing whether the current disturbance is a transient random fluctuation or a continuous deterioration trend, resulting in the digital twin model's insufficient ability to judge the direction of reliability risk evolution.

[0040] The Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time was estimated using rescaled range R / S analysis.

[0041] In estimating the Hurst exponent, the length of the sub-interval is set to an integer from 10 to W / 2, where W is the total number of samples within a preset time window before the acquisition time. The Hurst exponent at the acquisition time is determined by fitting a double log-linear regression. The Hurst exponent represents the strength of the long-term correlation of the corresponding multi-energy flow disturbance level sequence. A Hurst exponent greater than 0.5 indicates that the disturbance of the corresponding multi-energy flow disturbance level sequence has positive correlation persistence, that is, the disturbance will strengthen in the future. A Hurst exponent less than 0.5 indicates that the disturbance of the corresponding multi-energy flow disturbance level sequence has anti-persistence, that is, the disturbance will decay on its own in the future.

[0042] The collapse of multi-energy power plants is often triggered by rare, strongly coupled disturbances. Further quantification of the intensity and frequency of extreme disturbances is needed to improve the accuracy of digital twin models in assessing the probability of loss-making conditions. Rare, strongly coupled disturbances include extreme weather combined with equipment aging.

[0043] The peak over threshold (POT) algorithm is used to process the multi-energy flow disturbance level sequence at the time of acquisition. The initial value of the POT threshold is set to the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 1.5 times the standard deviation, so as to ensure that the threshold is higher than the normal fluctuation. If the Hurst exponent of the multi-energy flow disturbance level sequence at the time of acquisition is greater than the preset persistence threshold, the POT threshold is reassigned to the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 2 times the standard deviation. Under strong and persistent disturbances, the overall disturbance baseline rises, which will generate a large number of non-extreme high disturbance samples. The threshold is increased to filter out the pseudo-extreme points caused by the baseline rise, and to screen out the extreme disturbance samples with truly significant amplitude deviation from the baseline and continuous evolution capability, so as to capture more persistent extreme disturbances.

[0044] The persistence threshold is greater than 0.5 and less than 1. In this embodiment, the persistence threshold is set to 0.6.

[0045] Understandably, the peak over threshold algorithm characterizes the extreme event statistical features of a random process by fitting a generalized Pareto distribution to extreme data exceeding the POT threshold; a shape parameter greater than 0 indicates that the extreme value distribution features are heavy-tailed features, and the peak over threshold frequency represents the average number of times high-risk disturbances occur per hour.

[0046] The frequency exceeding the threshold at the acquisition time is used as the numerator, and the sum of the mean of the frequencies exceeding the threshold at all acquisition times before the acquisition time and the preset adjustment coefficient is used as the denominator. The value of the fraction is recorded as the relative frequency exceeding the threshold at the acquisition time. The maximum value between the shape parameter and the number 0 at the acquisition time is selected, and the product of the selected maximum value and the relative frequency exceeding the threshold at the acquisition time is recorded as the risk characteristic value at the acquisition time. The negative correlation processing result of the risk characteristic value at the acquisition time is recorded as the reliability assessment value at the acquisition time.

[0047] It is understood that negative correlation processing is applied to the risk characteristic value at the time of data collection, that is, to ensure that the risk characteristic value at the time of data collection is negatively correlated with the reliability assessment value at the time of data collection. It is understood that the negative correlation in this application refers to the relationship between the independent variable and the dependent variable, where the independent variable is the risk characteristic value at the time of data collection, and the dependent variable is the reliability assessment value at the time of data collection. The negative correlation means that the dependent variable decreases (increases) as the independent variable increases (increases), and can be an inverse relationship, a subtraction relationship, etc.

[0048] Preferably, as an embodiment of this application, the negative of the risk characteristic value at the time of acquisition is used as the exponent value, and the value of the exponential function with the natural constant as the base is recorded as the reliability assessment value at the time of acquisition.

[0049] The adjustment coefficient serves to prevent the denominator from being zero, and its specific value range is greater than or equal to... and less than or equal to In this embodiment, the specific value of the adjustment coefficient is... .

[0050] When the relative over-threshold frequency at the acquisition time is greater than 1, it indicates that the frequency of extreme disturbances is accelerating. The maximum value of the shape parameter and the number 0 at the acquisition time is used to ensure that the negative impact of heavy tail risk on reliability is considered. The smaller the relative over-threshold frequency and the maximum value of the shape parameter and the number 0 at the acquisition time, the less obvious the abnormal increase in the frequency of extreme events and the more controllable the tail risk is, and the more reliable the multi-energy station is.

[0051] At this point, the reliability assessment value at the time of data acquisition is obtained.

[0052] Step S004: Based on the reliability assessment value at the time of data acquisition, determine the convergence tolerance at the time of data acquisition, and conduct reliability simulation analysis of multi-energy power stations through the convergence tolerance.

[0053] This application introduces the sequential Monte Carlo simulation method for site reliability analysis. By gradually increasing the simulation sample size, the convergence of reliability is dynamically evaluated until a preset accuracy threshold is met, at which point the simulation terminates. However, in practical applications, the convergence criterion for the sequential Monte Carlo simulation method is usually set to a fixed absolute or relative tolerance. In multi-energy sites, the system operating state exhibits strong time-varying and non-stationary characteristics. A fixed tolerance would drive the algorithm to unnecessarily extract massive amounts of samples to reduce variance, resulting in a waste of computational resources.

[0054] To address the aforementioned issues, this application improves the dynamic convergence tolerance based on the reliability assessment value. Specifically, the convergence criterion in the sequential Monte Carlo simulation is constructed as a monotonically increasing function of the reliability assessment value. When the reliability assessment value is small, the tolerance is automatically narrowed to require higher simulation accuracy to capture extreme events. When the reliability assessment value is large, the original tolerance is maintained, and the simulation is terminated early while ensuring usable accuracy for engineering applications, thus saving computational resources.

[0055] The specific calculation method for the convergence tolerance at the acquisition time is as follows: take the difference between the number 1 and the reliability assessment value at the acquisition time as the numerator, take the sum of the product of the preset saturation coefficient and the reliability assessment value at the acquisition time and the number 1 as the denominator, take the product of the value of the fraction and the preset adjustment coefficient as the characteristic product at the acquisition time, and take the product of the difference between the number 1 and the characteristic product at the acquisition time and the preset initial tolerance as the convergence tolerance at the acquisition time.

[0056] Specifically, the saturation coefficient is used to prevent the tolerance from approaching zero excessively when the reliability assessment value is extremely small, causing the simulation to never terminate. The value of the saturation coefficient should be greater than or equal to 0.1 and less than or equal to 0.5. In this embodiment, the value of the saturation coefficient is 0.5. The adjustment coefficient is used to control the sensitivity of the convergence tolerance to changes in the reliability assessment value. The value of the adjustment coefficient should be greater than 0 and less than 1. In this embodiment, the value of the adjustment coefficient is 0.8. The value of the initial tolerance should be greater than 0.5 and less than 1.0. In this embodiment, the value of the initial tolerance is 0.8.

[0057] The smaller the reliability assessment value at the time of data acquisition, the greater the difference between the convergence tolerance at the time of acquisition and the initial tolerance, which can significantly improve the sampling accuracy under low reliability conditions. Therefore, the convergence tolerance at the time of acquisition can automatically increase the simulation input under low reliability conditions and actively reduce redundant calculations under high reliability conditions, realizing full-condition adaptive and efficient evaluation of reliability simulation for multi-energy power plants.

[0058] Specifically, the process of conducting reliability simulation analysis of multi-energy power plants through the aforementioned convergence tolerance includes: 1. Determine the variables to be sampled and their probability distributions. The variables to be sampled include the equipment failure rate of each energy station, as well as the environmental wind speed and irradiance parameters. Construct a comprehensive energy flow equation to describe the coupled steady-state operation of the electricity, heat and natural gas energy flows.

[0059] 2. Based on the digital twin model, initiate a sequential Monte Carlo simulation. In each iteration, generate a set of sampling variable parameters according to the probability distribution, substitute them into the comprehensive energy flow equation to solve, and calculate the system unload state under each iteration.

[0060] 3. During the simulation iteration process, the cumulative failure probability is continuously calculated as a reliability evaluation index, and the variance coefficient of the failure probability under the current iteration batch is calculated.

[0061] 4. When the obtained variance coefficient is less than or equal to the convergence tolerance, the simulation is determined to have reached the convergence criterion, the sequential Monte Carlo simulation iteration is terminated, and the current load failure probability is output as the multi-energy power station reliability simulation result.

[0062] This completes the reliability simulation of multi-energy power plants.

[0063] The multi-energy power station reliability simulation method specifically includes a physical power station layer, a sensor acquisition layer, a digital twin model layer, a reliability analysis layer, and a decision feedback layer.

[0064] The physical site layer includes, but is not limited to, wind turbines, photovoltaic arrays, energy storage batteries, gas turbines, and electric heating gas pipelines; the sensing and acquisition layer includes, but is not limited to, anemometers, irradiance meters, power sensors, and temperature sensors; the digital twin model layer is used to realize real-time high-fidelity mapping of the physical site; the reliability analysis layer is responsible for executing relevant algorithms; and the decision feedback layer is responsible for triggering early warnings and generating control commands such as energy storage scheduling, gas turbine power regulation, and load shedding.

[0065] Based on the same inventive concept as the above methods, embodiments of the present invention also provide a multi-energy power station reliability simulation system based on digital twins, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described multi-energy power station reliability simulation methods based on digital twins.

Claims

1. A reliability simulation method for multi-energy power plants based on digital twins, characterized in that, The method includes the following steps: The multi-energy station reliability simulation method includes a physical station layer, a sensor acquisition layer, a digital twin model layer, a reliability analysis layer, and a decision feedback layer. The simulation analysis process is as follows: the energy flow power of each energy station at each acquisition time and the environmental parameters of the multiple energy stations at each acquisition time are collected. The environmental parameter sequence and the energy flow power sequence of the energy station at the acquisition time are established. The energy flow includes electrical energy flow, thermal energy flow, and natural gas energy flow. The environmental parameters include ambient wind speed, irradiance, and temperature. Based on the empirical mode decomposition results of the environmental parameter sequence at the time of acquisition, and the coupling relationship between the power sequences of different energy flows at the time of acquisition, the multi-energy flow disturbance level at the time of acquisition is calculated, and the multi-energy flow disturbance level sequence at the time of acquisition is established. Based on the persistence and extreme value distribution characteristics of the multi-energy flow disturbance levels within the multi-energy flow disturbance level sequence at the time of acquisition, the reliability assessment value at the time of acquisition is calculated. Based on the reliability assessment value at the time of data acquisition, the convergence tolerance at the time of data acquisition is determined, and the reliability simulation analysis of multi-energy power stations is carried out through the convergence tolerance. The method for calculating the multi-energy flow disturbance level at the acquisition time is as follows: Based on the empirical mode decomposition results of the environmental parameter sequence at the time of acquisition, the proportion of high-frequency energy in the environmental parameter sequence is calculated. The coupling dependency difference at different time points is calculated based on the coupling relationship between the power sequences of different energy flows at the time of acquisition. The product of the arithmetic mean of the high-frequency energy proportions of all environmental parameter sequences at the acquisition time and the difference in coupling dependence is denoted as the multi-energy flow disturbance level at the acquisition time. The specific calculation method for the coupling dependency difference at the acquisition time is as follows: For the same acquisition time, the mutual information value of the power sequences of the two different energy streams is calculated, and the coefficient of variation of the mutual information value of the power sequences of the two different energy streams is recorded as the coupling dependency difference at the corresponding acquisition time.

2. The multi-energy power station reliability simulation method based on digital twins according to claim 1, characterized in that, The data in both the environmental parameter sequence and the energy flow power sequence are normalized.

3. The multi-energy power station reliability simulation method based on digital twins according to claim 1, characterized in that, The specific high-frequency energy proportion of the environmental parameter sequence is as follows: Empirical mode decomposition (EMD) is performed on the environmental parameter sequence at the acquisition time to obtain all IMF components. The ratio of the energy of the two highest-frequency IMF components to the total energy of all IMF components is recorded as the high-frequency energy ratio of the environmental parameter sequence.

4. The multi-energy power station reliability simulation method based on digital twins according to claim 1, characterized in that, The method for calculating the reliability assessment value at the acquisition time is as follows: Calculate the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time; Based on the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time and the values ​​of all multi-energy flow disturbance levels within the multi-energy flow disturbance level sequence, the POT threshold is set, and the shape parameter and over-threshold frequency at the acquisition time are calculated based on the multi-energy flow disturbance level sequence at the acquisition time. The reliability assessment value at the acquisition time is calculated based on the over-threshold frequency and shape parameters at the acquisition time.

5. The multi-energy power station reliability simulation method based on digital twins according to claim 4, characterized in that, The method for setting the POT threshold is as follows: If the Hurst exponent of the multi-energy flow disturbance level sequence at the acquisition time is less than or equal to the preset persistence threshold, the POT threshold is the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 1.5 times the standard deviation; otherwise, the POT threshold is the sum of the mean of all multi-energy flow disturbance levels in the multi-energy flow disturbance level sequence and 2 times the standard deviation.

6. The multi-energy power station reliability simulation method based on digital twins according to claim 4, characterized in that, The specific method for calculating the reliability assessment value at the acquisition time based on the over-threshold frequency and shape parameters at the acquisition time includes: The threshold frequency at the acquisition time is used as the numerator, and the sum of the mean of the threshold frequencies at all acquisition times before the acquisition time and the preset adjustment coefficient is used as the denominator. The value of the fraction is recorded as the relative threshold frequency at the acquisition time. The maximum value of the shape parameter and the number 0 at the acquisition time is selected, and the product of the selected maximum value and the relative over-threshold frequency at the acquisition time is recorded as the risk characteristic value at the acquisition time. The negative of the risk characteristic value at the time of data acquisition is used as the exponent value, and the value of the exponential function with the natural constant as the base is recorded as the reliability assessment value at the time of data acquisition.

7. The multi-energy power station reliability simulation method based on digital twins according to claim 1, characterized in that, The method for calculating the convergence tolerance at the acquisition time is as follows: The difference between the number 1 and the reliability assessment value at the acquisition time is used as the numerator. The sum of the product of the preset saturation coefficient and the reliability assessment value at the acquisition time and the number 1 is used as the denominator. The product of the value of the fraction and the preset adjustment coefficient is recorded as the characteristic product at the acquisition time. The product of the difference between the number 1 and the characteristic product at the acquisition time and the preset initial tolerance is recorded as the convergence tolerance at the acquisition time.

8. A multi-energy power station reliability simulation system based on digital twins, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as claimed in any one of claims 1-7.

Images