Power system operation reliability evaluation method and device and electronic equipment

Abstract

The application provides a power system operation reliability evaluation method and device and electronic equipment. The method comprises the following steps: obtaining a plurality of groups of historical data of a power system; for each group of historical data, sampling an element state of each element in the power system to obtain a plurality of system states, and calculating a historical reliability index of the group of historical data according to the plurality of system states; determining an operation scene to which each group of historical data belongs according to a preset classification condition; taking the plurality of groups of historical data and the corresponding historical reliability index in each operation scene as a data set, training a prediction model corresponding to the operation scene to obtain a plurality of trained prediction models, and constituting M-prediction models; and predicting real-time data of the power system obtained based on the M-prediction models to obtain a reliability index of the power system. The application can improve the evaluation speed of the power system operation reliability evaluation, and thus ensures the safety and stability of the power system operation.

Description

Technical Field

[0001] This application relates to the field of power system operation and maintenance technology, specifically to a method, apparatus, and electronic equipment for assessing the reliability of power system operation. Background Technology

[0002] In recent years, the proportion of new energy power generation in the power system has increased rapidly. However, new energy power generation is random and uncertain. In order to ensure the safe and stable operation of the power system, it is necessary to assess the reliability of the power system operation.

[0003] Conventional methods for assessing the reliability of power system operation are computationally slow, as the computation time depends on both the scale of the power system and the required convergence accuracy. This often results in slow computation speeds that fail to meet the demands for rapid assessment. Therefore, improving the speed of power system reliability assessments and ensuring the safety and stability of power system operation are pressing issues that need to be addressed in this field. Summary of the Invention

[0004] In view of this, embodiments of this application provide a method, apparatus, and electronic device for assessing the reliability of power system operation, in order to solve the technical problem that conventional methods for assessing the reliability of power system operation are slow in calculation and cannot meet the requirements for rapid assessment.

[0005] In a first aspect, embodiments of this application provide a method for assessing the operational reliability of a power system, comprising: acquiring multiple sets of historical data of the power system; for each set of historical data, sampling the component state of each element in the power system using a non-sequential Monte Carlo sampling method to obtain multiple system states, calculating the load reduction probability and optimal load reduction amount for each system state, and calculating the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount for the multiple system states; determining the operating scenario to which each set of historical data belongs according to preset classification conditions; using multiple sets of historical data and corresponding historical reliability indices under each operating scenario as a dataset, training a prediction model corresponding to the operating scenario to obtain multiple trained prediction models, constituting an M-prediction model; acquiring real-time data of the power system, predicting the real-time data based on the M-prediction model, and obtaining the reliability index of the power system, so as to assess the power system based on the reliability index.

[0006] In one possible implementation of the first aspect, the prediction model is a backpropagation neural network; there are multiple running scenarios; multiple sets of historical data and corresponding historical reliability indicators under each running scenario are used as a dataset, and the prediction model corresponding to the running scenario is trained to obtain multiple trained prediction models, which constitute an M-prediction model, including: using historical data under each running scenario as model input and historical reliability indicators as model output, training the prediction model corresponding to the running scenario based on an adaptive algorithm to obtain the trained prediction model corresponding to the running scenario; and constructing an M-prediction model from the multiple trained prediction models corresponding to multiple running scenarios.

[0007] In one possible implementation of the first aspect, the training process of the prediction model corresponding to each operating scenario includes: enumerating the number of hidden layers and the number of hidden elements of the prediction model according to the preset maximum number of hidden layers and the preset maximum number of hidden elements; training and testing the prediction model based on the dataset of the operating scenario corresponding to the prediction model according to the currently enumerated number of hidden layers and the number of hidden elements to obtain the error of the current reliability index; after all the number of hidden layers and the number of hidden elements have been enumerated, obtaining the errors of multiple reliability indices, determining the number of hidden layers and the number of hidden elements corresponding to the smallest error of the reliability index as the final number of hidden layers and the number of hidden elements of the prediction model, and obtaining the trained prediction model.

[0008] In one possible implementation of the first aspect, each system state is an array of element states for each element; each element state is a first state or a second state, wherein the first state represents a faulty element state and the second state represents a non-faulty element state; multiple system states are obtained by sampling the element states of each element in the power system using a non-sequential Monte Carlo sampling method, including: determining the element state of each element using a non-sequential Monte Carlo sampling method based on the unavailability rate of each element in the power system, determining the system state based on the element state of each element; and obtaining multiple system states by sampling the element states of each element in the power system multiple times.

[0009] In one possible implementation of the first aspect, calculating the load reduction probability and optimal load reduction amount for each system state includes: calculating the load reduction probability for each system state based on the element state of each element corresponding to each system state; constructing an optimal load shedding model; and calculating the optimal load reduction amount for each system state based on the optimal load shedding model.

[0010] In one possible implementation of the first aspect, the reliability index includes the load shedding probability and the expected load shedding amount; the historical reliability index of the set of historical data is calculated based on the load shedding probability and the optimal load shedding amount of multiple system states, including: summing the load shedding probabilities of multiple system states to obtain the historical load shedding probability of the set of historical data; and calculating the historical expected load shedding amount of the set of historical data by weighted summing the optimal load shedding amounts of multiple system states based on the load shedding probabilities of multiple system states.

[0011] In one possible implementation of the first aspect, the real-time data includes real-time wind power output, real-time photovoltaic power output, and real-time load; acquiring real-time data of the power system and predicting the real-time data based on the M-prediction model to obtain the reliability index of the power system includes: acquiring real-time data of the power system, determining the operating scenario to which the real-time data belongs according to preset classification conditions; and predicting the real-time data based on the prediction model corresponding to the operating scenario in the M-prediction model to obtain the reliability index of the power system.

[0012] Secondly, embodiments of this application provide a power system operation reliability assessment device, comprising:

[0013] The acquisition module is used to acquire multiple sets of historical data from the power system.

[0014] The calculation module is used to sample the component state of each element in the power system for each set of historical data using non-sequential Monte Carlo sampling to obtain multiple system states, calculate the load reduction probability and optimal load reduction amount for each system state, and calculate the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount of multiple system states.

[0015] The module is used to determine the operating scenario to which each set of historical data belongs according to preset classification conditions; multiple sets of historical data and corresponding historical reliability indicators under each operating scenario are used as datasets to train the prediction model corresponding to the operating scenario, resulting in multiple trained prediction models, which constitute the M-prediction model.

[0016] The prediction module is used to acquire real-time data of the power system, predict the real-time data based on the M-prediction model, and obtain the reliability index of the power system, so as to evaluate the power system based on the reliability index.

[0017] Thirdly, embodiments of this application provide an electronic device, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the power system operation reliability assessment method as described in any of the first aspects.

[0018] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power system operation reliability assessment method as described in any of the first aspects.

[0019] Fifthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to execute the power system operation reliability assessment method described in any of the first aspects above.

[0020] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0021] The power system operation reliability assessment method, apparatus, and electronic equipment provided in this application acquire multiple sets of historical data of the power system, calculate the historical reliability index of each set of historical data, determine the operation scenario to which each set of historical data belongs according to preset classification conditions, use multiple sets of historical data and corresponding historical reliability indices under each operation scenario as a dataset, train the prediction model corresponding to the operation scenario to obtain multiple trained prediction models, forming an M-prediction model, and predict the acquired real-time data of the power system based on the M-prediction model to obtain the reliability index of the power system. By training different prediction models corresponding to different operation scenarios and determining the operation scenario to which the acquired real-time data of the power system belongs, and using the prediction model corresponding to the operation scenario to predict the real-time data, the assessment speed of power system operation reliability assessment can be improved, thereby ensuring the safety and stability of power system operation.

[0022] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this specification. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic flowchart of a power system operation reliability assessment method provided in an embodiment of this application;

[0025] Figure 2 This is a flowchart illustrating a power system operation reliability assessment method provided in another embodiment of this application;

[0026] Figure 3 This is a flowchart illustrating a power system operation reliability assessment method provided in another embodiment of this application;

[0027] Figure 4 This is a flowchart of a power system operation reliability assessment method provided in an embodiment of this application;

[0028] Figure 5 This is a comparison chart of the unload expectation error of different neural network models provided in an embodiment of this application;

[0029] Figure 6 This is a comparison diagram of the error probability density of the unload expectation of different neural network models provided in an embodiment of this application;

[0030] Figure 7 This is a comparison chart of the expected load loss error of the M-prediction model under different classification criteria provided in an embodiment of this application;

[0031] Figure 8 This is a schematic diagram of the structure of a power system operation reliability assessment device provided in one embodiment of this application;

[0032] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0033] The present application will be described more clearly below with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the function of the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0034] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0035] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0036] In the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0037] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0038] Furthermore, the term "multiple" mentioned in the embodiments of this application should be interpreted as two or more.

[0039] Figure 1 This is a flowchart illustrating a power system operation reliability assessment method provided in an embodiment of this application. Figure 1 As shown, the method in the embodiments of this application may include:

[0040] Step 101: Obtain multiple sets of historical data from the power system.

[0041] Optionally, the historical data mentioned above includes historical wind power output, historical photovoltaic power output, and historical load.

[0042] Step 102: For each set of historical data, sample the component state of each element in the power system using non-sequential Monte Carlo sampling to obtain multiple system states. Calculate the load reduction probability and optimal load reduction amount for each system state. Calculate the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount for multiple system states.

[0043] The components in the power system include multiple wind turbines, multiple photovoltaic units, and multiple power lines. Historical reliability indicators include historical load shedding probability and historical load shedding expectation.

[0044] In one possible implementation, see Figure 2 Step 102 may specifically include:

[0045] Step 1021: For each set of historical data, based on the unavailability rate of each component in the power system, determine the component state of each component using non-sequential Monte Carlo sampling, and determine the system state based on the component state of each component.

[0046] The unavailability rate of a component is its failure probability.

[0047] For example, in each photovoltaic (PV) unit, a failure of a PV module will only cause the shutdown of the PV string to which it belongs; a failure of an inverter will cause the shutdown of the PV power generation unit to which it belongs; and a failure of an isolation transformer will cause the entire PV power station to shut down. Therefore, PV units can be divided into three layers: a DC layer, an inverter layer, and a grid-connected layer. A three-layer model can be constructed, and the failure probability of PV units in the power system can be calculated using the module string-parallel reliability parameter calculation method.

[0048] in,

[0049]

[0050] In the formula, the failure rates of photovoltaic modules and fuses are λ and λ, respectively. c and λ f The mean time to repair the faults were r c and r f n is the number of photovoltaic modules connected in series in the photovoltaic string; λ1, r1, and U1 represent the failure rate, fault repair time, and fault probability of the DC layer, respectively; the failure rates of the DC circuit breaker, inverter, and AC circuit breaker are λ1, r1, and U1, respectively. dcb , λ inv and λ acb The mean time to repair the faults were r dcb r inv and r acb λ², r², and U² represent the failure rate, fault repair time, and fault probability of the inverter layer, respectively; the failure rates of the AC disconnect switch, grid-connected circuit breaker, and step-up transformer are λ², r², and U², respectively. acs , λ gcb and λ bt The mean time to repair the faults were r acs r gcb and r bt λ3, r3, and U3 represent the failure rate, fault repair time, and fault probability of the inverter layer, respectively.

[0051] The formula for the failure probability of a photovoltaic unit is:

[0052]

[0053] In the formula, U(I,J,K) represents the failure probability of the photovoltaic unit, I represents the failure order of each component in the DC layer, J represents the failure order of each component in the inverter layer, and K represents the failure order of each component in the grid-connected layer; h represents the number of photovoltaic inverters connected in parallel in the inverter layer, m represents the number of photovoltaic strings, and C represents the permutation and combination formula. The failure probability of each photovoltaic unit, i.e., the unavailability rate of each photovoltaic unit, can be calculated using the above formula.

[0054] Optionally, the unavailability rate of wind turbines and power lines can be set as needed, without specific restrictions here.

[0055] For example, non-sequential Monte Carlo sampling is based on the premise that a system state is a combination of all component states, and each component state can be determined by sampling the probability of that component appearing in that state. The formula for the component state is:

[0056]

[0057] In the formula, S x U represents the component state of component x. x Q is a random number drawn from component x. x Let be the unavailability rate of component x. For each photovoltaic unit, the unavailability rate is calculated using the photovoltaic unit failure probability formula. For each wind turbine and each line, the unavailability rate can be set as needed.

[0058] Each component is sampled to determine its component state. The component states of all components in the power system constitute the system state of the power system. From the formula for component states, each component's component state is either a first state or a second state. The first state represents a fault state, which can be represented by 0, while the second state represents a non-fault state, which can be represented by 1. Therefore, each system state is an array of component states for each component. For example, the system state can be represented as 010011101111…0111.

[0059] Step 1022: Sample the state of each component in the power system multiple times to obtain multiple system states.

[0060] For example, for each set of historical data, the component state of each element in the power system is sampled 100 times, resulting in 100 system states.

[0061] Step 1023: Calculate the load reduction probability for each system state based on the component state of each component corresponding to each system state.

[0062] Optionally, the number of components in each system state that are in state 1 is calculated. The load reduction probability for that system state is calculated by dividing the number of components in state 1 by the total number of components. In other words, each system state is an array of component states for each component. The number of zero elements in this array is calculated, and the number of zero elements is divided by the total number of elements to obtain the load reduction probability for that system state. For example, if a power system contains 100,000 components, then the array corresponding to a certain system state will contain 100,000 elements, of which 40,000 will be zero elements. Therefore, the load reduction probability for that system state is 0.4.

[0063] Step 1024: Construct the optimal load shedding model and calculate the optimal load shedding amount for each system state based on the optimal load shedding model.

[0064] The optimal load shedding model can be expressed as:

[0065]

[0066] The constraints are:

[0067]

[0068] P Gimin ≤P Gi ≤P Gimax i∈N G

[0069] 0≤P Ci ≤P Di i∈N D

[0070] |P k |≤P kmax k∈N B

[0071] In the formula, LC s P represents the optimal load shedding amount for system state s, Ω is the number of nodes in the power system, each node has one generator unit, which is either a wind turbine or a photovoltaic generator unit, and every two adjacent nodes are connected by a line; Di and P Ci These are the initial load and load shedding values ​​at the i-th node, respectively; N G N B N D Each is a set of generating units, lines, and nodes; P Gimin P Gi P Gimax These represent the minimum output, actual output, and maximum output of the i-th generator unit, respectively; P k and P kmaxThese are the actual and maximum power flow values ​​for the k-th line, respectively.

[0072] Step 1025: Sum the load reduction probabilities of multiple system states to obtain the historical load loss probability of this set of historical data.

[0073] Step 1026: Based on the load reduction probabilities of multiple system states, perform a weighted sum calculation on the optimal load reduction amount for multiple system states to obtain the expected historical load loss amount of this set of historical data.

[0074] The formula for the historical load failure probability is as follows:

[0075]

[0076] The formula for the expected amount of historical load shedding is:

[0077]

[0078] In the formula, LOLP represents the historical load loss probability, EENS represents the expected historical load loss, D represents the total state space generated by non-sequential Monte Carlo sampling, which is all system states, and P s LC represents the load reduction probability of system state s. s The value represents the optimal load shedding amount for system state s, and T represents the evaluation period. For example, the evaluation period can be set to six months or three months as needed.

[0079] It is important to note that for each set of historical data, the variance coefficient of the reliability index is calculated, and it is determined whether the variance coefficient meets the preset convergence condition. If the variance coefficient meets the preset convergence condition, step 103 is executed; if the variance coefficient does not meet the preset convergence condition, it indicates that the number of samplings is insufficient. In this case, step 1022 is executed again to obtain multiple new system states. These multiple new system states are combined with the previous multiple system states to form multiple system states, and steps 1023 to 1026 are executed until the variance coefficient of the reliability index of the set of historical data meets the preset convergence condition.

[0080] For each set of historical data, perform steps 1021 to 1026 as described above to obtain the historical reliability index of each set of historical data, that is, to obtain the historical load failure probability and the expected historical load failure amount of each set of historical data.

[0081] Step 103: Determine the operating scenario to which each group of historical data belongs according to the preset classification conditions.

[0082] There are multiple operating scenarios.

[0083] Optionally, as mentioned above, historical data includes historical wind power output, historical solar power output, and historical load. Therefore, preset classification conditions can include wind speed conditions, solar intensity conditions, and load conditions. Classification can be based on mean or quantile standards. For example, assuming a maximum wind speed of 60 km / h, according to the mean classification standard, wind speed conditions can be categorized as follows: wind speed less than or equal to 30 km / h is category one, and wind speed greater than 30 km / h is category two. That is, based on the wind speed conditions and the corresponding wind speeds for historical wind power output, historical wind power output is divided into two categories. Specifically, the classification standards and preset classification conditions can be set as needed.

[0084] Similarly, if wind speed conditions can classify historical wind power output into category X, solar intensity conditions can classify historical photovoltaic output into category Y, and load conditions can classify historical load into category Z, then there are a total of X×Y×Z operating scenarios.

[0085] For each set of historical data, the operating scenario to which the historical data belongs is determined based on wind speed conditions, light intensity conditions, and load conditions. Thus, each operating scenario includes multiple sets of historical data.

[0086] Step 104: Take multiple sets of historical data and corresponding historical reliability indicators for each operating scenario as a dataset, train the prediction model corresponding to the operating scenario, and obtain multiple trained prediction models to form an M-prediction model.

[0087] In one possible implementation, step 104 may specifically include:

[0088] Step 1041: Using historical data for each operating scenario as model input and historical reliability index as model output, train the prediction model corresponding to the operating scenario based on the adaptive algorithm to obtain the trained prediction model corresponding to the operating scenario.

[0089] Optionally, each of the above-mentioned operating scenarios corresponds to multiple sets of historical data belonging to that operating scenario, and a corresponding prediction model, which is a Back Propagation Neural Network (BPNN). By determining the operating scenario to which each set of historical data belongs, and using the multiple sets of historical data corresponding to each operating scenario as a dataset to train the prediction model corresponding to that operating scenario, the features of the historical data can be extracted better, resulting in higher accuracy for the prediction model corresponding to each operating scenario.

[0090] For example, see Figure 3 The training process of the prediction model for each running scenario can specifically include:

[0091] S1. Enumerate the number of hidden layers and the number of hidden elements of the prediction model according to the preset maximum number of hidden layers and the preset maximum number of hidden elements.

[0092] S2. Based on the currently enumerated number of hidden layers and hidden elements, train and test the prediction model using the dataset in the corresponding running scenario to obtain the error of the current reliability index.

[0093] S3. After enumerating all hidden layers and hidden elements, the errors of multiple reliability indicators are obtained. The hidden layers and hidden elements corresponding to the smallest reliability indicator error are determined as the final hidden layers and hidden elements of the prediction model, thus obtaining the trained prediction model.

[0094] Specifically, the dataset in this operating scenario is divided into a training set and a test set according to a preset ratio. The prediction model is trained using the training set and tested using the test set. The error of the reliability index is calculated during the testing process.

[0095] Optionally, based on the preset maximum number of hidden layers and the maximum number of hidden elements of the prediction model, the number of hidden layers and the number of hidden elements of the prediction model are enumerated, that is, the number of hidden layers and the number of hidden elements of the prediction model are changed, the prediction model is trained and tested, and the number of hidden layers and the number of hidden elements corresponding to the minimum reliability index error are determined as the final number of hidden layers and the number of hidden elements of the prediction model, thus obtaining the trained prediction model.

[0096] Step 1042: Combine multiple trained prediction models corresponding to multiple running scenarios into an M-prediction model.

[0097] Optionally, each running scenario corresponds to a trained prediction model, and all trained prediction models are combined to form an M-prediction model (M-BPNN) as the final prediction model.

[0098] Step 105: Obtain real-time data of the power system, predict the real-time data based on the M-prediction model, obtain the reliability index of the power system, and evaluate the power system based on the reliability index.

[0099] The real-time data includes real-time wind power output, real-time photovoltaic power output, and real-time load.

[0100] In one possible implementation, step 105 may specifically include:

[0101] Acquire real-time data from the power system and determine the operating scenario to which the real-time data belongs based on preset classification conditions.

[0102] Based on the M-prediction model, the prediction model corresponding to this operating scenario predicts real-time data to obtain the reliability index of the power system.

[0103] Optional, see Figure 4 The real-time wind power output is classified according to wind speed conditions, the real-time photovoltaic power output is classified according to light intensity conditions, and the real-time load is classified according to load conditions to determine the operating scenario to which the real-time data belongs. Based on the above operating scenarios, the corresponding prediction model is determined. The real-time data is then predicted based on the prediction model to obtain the reliability index of the power system, namely the power system load failure probability and the expected load failure amount, so as to evaluate the operational reliability of the power system based on the above power system reliability index.

[0104] Determining the operating scenario to which the acquired real-time power system data belongs, and using the prediction model corresponding to that operating scenario to predict the real-time data, can improve the calculation speed of power system operation reliability assessment and shorten the assessment time. At the same time, using the prediction model corresponding to the operating scenario can improve the accuracy of the predicted reliability indicators, that is, improve the effectiveness of the assessment results.

[0105] This application provides a method for assessing the reliability of power system operation. It acquires multiple sets of historical data from the power system, calculates historical reliability indices for each set, determines the operating scenario to which each set of historical data belongs based on preset classification conditions, and uses multiple sets of historical data and corresponding historical reliability indices under each operating scenario as a dataset. A prediction model corresponding to that operating scenario is trained to obtain multiple trained prediction models, forming an M-prediction model. Based on the M-prediction model, the acquired real-time data of the power system is predicted to obtain the power system's reliability index. By training different prediction models corresponding to different operating scenarios and determining the operating scenario to which the acquired real-time data of the power system belongs, and using the prediction model corresponding to that operating scenario to predict the real-time data, the assessment speed of power system operation reliability can be improved, thereby ensuring the safety and stability of power system operation.

[0106] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0107] To verify the evaluation speed and accuracy of the power system operation reliability assessment method provided in the embodiments of this application, conventional power system operation reliability assessment methods, such as the traditional Monte Carlo simulation method, are compared with the power system operation reliability assessment method provided in the embodiments of this application.

[0108] Specifically, based on the acquired real-time data of the power system, the reliability indicators of the power system are calculated using both the traditional Monte Carlo simulation (MC) method and the power system operation reliability assessment method provided in this application embodiment. Specifically, the power system's load failure probability and expected load failure amount, as well as the assessment time, are calculated. The reliability assessment results of the two methods are shown in Table 1.

[0109] Table 1

[0110]

[0111] As shown in Table 1, the evaluation time of the power system operation reliability assessment method provided in this application embodiment is 0.9101 seconds, which meets the time requirement for rapid evaluation and is much shorter than the evaluation time of 21.8339 seconds for the traditional Monte Carlo simulation method. This indicates that the power system operation reliability assessment method provided in this application embodiment can improve the evaluation speed of power system operation reliability assessment. Meanwhile, the load shedding probability error and the expected load shedding error of the power system operation reliability assessment method provided in this application embodiment are 0.1356%, indicating that the reliability assessment results are effective and accurate.

[0112] Furthermore, to verify the effectiveness of the M-Prediction Model (M-BPNN), three neural network models—Radial Basis Function (RBF) neural network, General Regression Neural Network (GRNN), and the M-Prediction Model (M-BPNN) provided in this application—were used to predict the reliability indices of the power system. The load shedding probability and expected load shedding amount were calculated under 100 sets of sampled power system data (referred to as 100 system operating states). The overall errors of the load shedding probability and expected load shedding amount for the three neural network models are shown in Table 2. A comparison of the errors in the expected load shedding amount for the three neural network models is also provided. Figure 5 As shown, the comparison of the error probability density of the unload expectation of the above three neural network models is as follows: Figure 6 As shown.

[0113] Table 2

[0114]

[0115] As shown in Table 2, the M-prediction model provided in this application has the smallest overall error in both the probability of load failure and the expected amount of load failure, at 0.5924% and 0.4346%, respectively. Figure 5 and Figure 6It can be seen that the error of the expected load shedding amount of the M-prediction model provided in this application embodiment is small, within ±5%. At the same time, the error probability density of the expected load shedding amount of the M-prediction model provided in this application embodiment is concentrated around 0, indicating that only small errors will occur in most cases.

[0116] Furthermore, to verify the sensitivity of the M-prediction model (M-BPNN), historical data was classified using different classification criteria (lower tertiary, median, and upper tertiary) to train the M-prediction model provided in this application, resulting in three M-prediction models under different classification criteria. The M-prediction models under different classification criteria were used to predict the reliability indicators of the power system, and the probability of load shedding and the expected load shedding were calculated for 100 system operating states. The overall error of the probability of load shedding and the overall error of the expected load shedding of the M-prediction model under the above different classification criteria are shown in Table 3. A comparison of the errors of the expected load shedding of the M-prediction model under the above different classification criteria is also provided. Figure 7 As shown.

[0117] Table 3

[0118]

[0119] From Table 3 and Figure 7 It is evident that different classification criteria lead to significant differences in the M-prediction model. Under the low tertiary classification criterion, the overall error of the failure probability and the overall error of the failure expectation of the M-prediction model are both large. This may be due to two reasons: first, some categories under this classification criterion lack training data, resulting in larger prediction errors; second, under this classification criterion, data with large differences in reliability index weights are classified into the same category, leading to inaccurate prediction results.

[0120] Figure 8 This is a schematic diagram of the structure of a power system operation reliability assessment device provided in one embodiment of this application. Figure 8 As shown, the power system operation reliability assessment device provided in this embodiment may include: an acquisition module 201, a calculation module 202, a construction module 203, and a prediction module 204.

[0121] The acquisition module 201 is used to acquire multiple sets of historical data from the power system.

[0122] The calculation module 202 is used to sample the component state of each element in the power system for each set of historical data using the non-sequential Monte Carlo sampling method to obtain multiple system states, calculate the load reduction probability and optimal load reduction amount for each system state, and calculate the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount of multiple system states.

[0123] The construction module 203 is used to determine the operating scenario to which each group of historical data belongs according to the preset classification conditions; the multiple groups of historical data and corresponding historical reliability indicators under each operating scenario are used as a dataset to train the prediction model corresponding to the operating scenario, thereby obtaining multiple trained prediction models, which constitute the M-prediction model.

[0124] The prediction module 204 is used to acquire real-time data of the power system, predict the real-time data based on the M-prediction model, and obtain the reliability index of the power system so as to evaluate the power system based on the reliability index.

[0125] Optionally, the prediction model is a backpropagation neural network; there are multiple running scenarios, and the construction module 203 is specifically used to: take the historical data of each running scenario as the model input, the historical reliability index as the model output, train the prediction model corresponding to the running scenario based on the adaptive algorithm, and obtain the trained prediction model corresponding to the running scenario; and construct an M-prediction model from the multiple trained prediction models corresponding to multiple running scenarios.

[0126] Optionally, the construction module 203 is further specifically used for: enumerating the number of hidden layers and the number of hidden elements of the prediction model according to the preset maximum number of hidden layers and the preset maximum number of hidden elements; training and testing the prediction model based on the dataset of the running scenario corresponding to the prediction model according to the currently enumerated number of hidden layers and the number of hidden elements, and obtaining the error of the current reliability index; after all the number of hidden layers and the number of hidden elements have been enumerated, obtaining the errors of multiple reliability indices, determining the number of hidden layers and the number of hidden elements corresponding to the smallest error of the reliability index as the final number of hidden layers and the number of hidden elements of the prediction model, and obtaining the trained prediction model.

[0127] Optionally, each system state is an array of element states for each component; each element state is either a first state or a second state, where the first state represents a faulty element state and the second state represents a non-faulty element state; the calculation module 202 is specifically used to: determine the element state of each component based on the unavailability rate of each component in the power system using a non-sequential Monte Carlo sampling method, determine the system state based on the element state of each component, and perform multiple samplings on the element state of each component in the power system to obtain multiple system states.

[0128] Optionally, the calculation module 202 is also specifically used to: calculate the load reduction probability of each system state based on the component state of each component corresponding to each system state; construct an optimal load shedding model, and calculate the optimal load reduction amount for each system state based on the optimal load shedding model.

[0129] Optionally, the reliability indicators include the load shedding probability and the expected load shedding amount; the calculation module 202 is also specifically used to: sum the load shedding probabilities of multiple system states to obtain the historical load shedding probability of the set of historical data; and calculate the weighted sum of the optimal load shedding amount of multiple system states based on the load shedding probabilities of multiple system states to obtain the historical load shedding amount of the set of historical data.

[0130] Optionally, the real-time data includes real-time wind power output, real-time photovoltaic power output, and real-time load; the prediction module 204 is specifically used to: acquire real-time data of the power system, determine the operating scenario to which the real-time data belongs based on preset classification conditions; and predict the real-time data based on the prediction model corresponding to the operating scenario in the M-prediction model to obtain the reliability index of the power system.

[0131] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0132] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. For example... Figure 9 As shown, the electronic device 300 of this embodiment includes a processor 310 and a memory 320, wherein the memory 320 stores a computer program 321 that can run on the processor 310. When the processor 310 executes the computer program 321, it implements the steps in any of the above-described method embodiments, for example... Figure 1 Steps 101 to 105 are shown. Alternatively, when processor 310 executes computer program 321, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 8 The functions of modules 201 to 204 are shown.

[0133] For example, computer program 321 may be divided into one or more modules / units, one or more of which are stored in memory 320 and executed by processor 310 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 321 in electronic device 300.

[0134] Those skilled in the art will understand that Figure 9 This is merely an example of an electronic device and does not constitute a limitation on the electronic device. It may include more or fewer components than shown, or combinations of certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0135] The processor 310 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0136] The memory 320 can be an internal storage unit of the electronic device, such as a hard drive or memory, or an external storage device, such as a plug-in hard drive, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card. The memory 320 can also include both internal and external storage units. The memory 320 is used to store computer programs and other programs and data required by the electronic device. The memory 320 can also be used to temporarily store data that has been output or will be output.

[0137] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0138] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0139] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0140] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0141] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0142] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0143] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0144] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for assessing the operational reliability of a power system, characterized in that, include: Acquire multiple sets of historical data of the power system; for each set of historical data, sample the element state of each component in the power system using non-sequential Monte Carlo sampling method to obtain multiple system states, calculate the load reduction probability and optimal load reduction amount for each system state, and calculate the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount of the multiple system states. According to the preset classification conditions, the operating scenario to which each group of historical data belongs is determined, wherein the preset classification conditions include wind speed conditions, light intensity conditions and load conditions; multiple groups of historical data and corresponding historical reliability indicators under each operating scenario are used as datasets, and the prediction model corresponding to the operating scenario is trained to obtain multiple trained prediction models, which constitute an M-prediction model. Real-time data of the power system is acquired, and the real-time data is predicted based on the M-prediction model to obtain the reliability index of the power system, so as to evaluate the power system according to the reliability index. Each system state is an array of the element states of each component; each element state is either a first state or a second state, wherein the first state represents a faulty element state and the second state represents a non-faulty element state; the number of elements in each system state that are in the first state is calculated, and the ratio of the number of elements in the first state to the total number of elements is used as the load reduction probability of the system state. The reliability metrics include the probability of load failure and the expected load failure. The calculation of historical reliability indices for this set of historical data based on the load reduction probability and optimal load reduction amount of the multiple system states includes: The load reduction probabilities of the multiple system states are summed to obtain the historical load loss probability of this set of historical data; Based on the load reduction probabilities of the multiple system states, the optimal load reduction amount for the multiple system states is weighted and calculated to obtain the expected historical load loss amount of the set of historical data.

2. The power system operation reliability assessment method according to claim 1, characterized in that, The prediction model is a backpropagation neural network; there are multiple operating scenarios; The process involves using multiple sets of historical data and corresponding historical reliability indicators for each operational scenario as a dataset to train a prediction model for that operational scenario, resulting in multiple trained prediction models that constitute an M-prediction model, including: Using historical data for each operational scenario as model input and historical reliability index as model output, the prediction model corresponding to the operational scenario is trained based on an adaptive algorithm to obtain the trained prediction model corresponding to the operational scenario. The M-prediction model is formed by combining multiple trained prediction models corresponding to multiple operating scenarios.

3. The power system operation reliability assessment method according to claim 1, characterized in that, The training process of the prediction model for each running scenario includes: Based on the preset maximum number of hidden layers and the preset maximum number of hidden elements of the prediction model, the number of hidden layers and the number of hidden elements of the prediction model are enumerated. Based on the currently enumerated number of hidden layers and hidden elements, the prediction model is trained and tested on the dataset of the running scenario corresponding to the prediction model, and the error of the current reliability index is obtained. After all hidden layers and hidden elements have been enumerated, the errors of multiple reliability indicators are obtained. The hidden layers and hidden elements corresponding to the smallest reliability indicator error are determined as the final hidden layers and hidden elements of the prediction model, thus obtaining the trained prediction model.

4. The power system operation reliability assessment method according to claim 1, characterized in that, The non-sequential Monte Carlo sampling method is used to sample the element state of each component in the power system, resulting in multiple system states, including: Based on the unavailability rate of each component in the power system, the component state of each component is determined by non-sequential Monte Carlo sampling, and the system state is determined based on the component state of each component. The component state of each element in the power system is sampled multiple times to obtain multiple system states.

5. The power system operation reliability assessment method according to claim 1, characterized in that, Calculating the optimal load shedding for each of the system states includes: Construct an optimal load shedding model, and calculate the optimal load reduction amount for each system state based on the optimal load shedding model.

6. The power system operation reliability assessment method according to any one of claims 1 to 5, characterized in that, The real-time data includes real-time wind power output, real-time photovoltaic power output, and real-time load. The process of acquiring real-time data of the power system, predicting the real-time data based on the M-prediction model, and obtaining the reliability index of the power system includes: Acquire real-time data of the power system and determine the operating scenario to which the real-time data belongs based on the preset classification conditions; Based on the M-prediction model, the prediction model corresponding to the operating scenario predicts the real-time data to obtain the reliability index of the power system.

7. A power system operation reliability assessment device, characterized in that, include: The acquisition module is used to acquire multiple sets of historical data from the power system. The calculation module is used to sample the element state of each component in the power system for each set of historical data using non-sequential Monte Carlo sampling method to obtain multiple system states, calculate the load reduction probability and optimal load reduction amount for each system state, and calculate the historical reliability index of the set of historical data based on the load reduction probability and optimal load reduction amount of the multiple system states. The construction module is used to determine the operating scenario to which each group of historical data belongs according to preset classification conditions, wherein the preset classification conditions include wind speed conditions, light intensity conditions and load conditions; multiple groups of historical data and corresponding historical reliability indicators under each operating scenario are used as datasets to train the prediction model corresponding to the operating scenario, thereby obtaining multiple trained prediction models, which constitute an M-prediction model. The prediction module is used to acquire real-time data of the power system, predict the real-time data based on the M-prediction model, obtain the reliability index of the power system, and evaluate the power system based on the reliability index. Each system state is an array of the component states of each element; each component state is a first state or a second state, wherein the first state represents a faulty component state, and the second state represents a non-faulty component state; the reliability indicators include the probability of load failure and the expected load failure amount. The calculation module is further configured to calculate the number of components in the first state in each system state, divide the number of components in the first state by the total number of components, and use the ratio as the load reduction probability of the system state; and sum the load reduction probabilities of the multiple system states to obtain the historical load loss probability of the set of historical data; and calculate the optimal load reduction amount of the multiple system states based on the load reduction probabilities of the multiple system states to obtain the expected historical load loss amount of the set of historical data.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the power system operation reliability assessment method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the power system operation reliability assessment method as described in any one of claims 1 to 6.

Images