Distribution area power supply reliability evaluation method and device based on neural network model

Abstract

The invention discloses a zone area power supply reliability evaluation method and device based on a neural network model, and the method comprises the steps: determining a target zone area in a power distribution network when a reliability evaluation request is received; acquiring historical power supply data of the target station area, historical action data of the switch equipment, topological line structure data and historical fault data; inputting the historical power supply data, the historical action data, the topological line structure data and the historical fault data into a reliability scoring model to obtain a power supply reliability score of the target transformer area; the power supply reliability grade of the target area is determined based on the reliability grade, manual data collection and scoring are not needed, manpower is saved, the subjective influence of manual scoring can be eliminated, and the power supply reliability grade of the target area can be accurately and objectively obtained, so that the reliability grade is accurately determined; and an accurate reference is provided for system maintenance of a target station area.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply of distribution network, and in particular to a method and device for evaluating the reliability of power supply in a distribution area based on a neural network model. Background Art

[0002] The operational reliability of the power system is related to the national economy, people's livelihood and national security. Accurately assessing the operational reliability of the power system can provide targeted guidance for system maintenance.

[0003] In the prior art, the reliability of power supply in the distribution network is mainly evaluated by collecting fault data of the distribution network, such as the number of faults, the number of power outages, the duration of power outages, the duration of power restoration, etc., and manually scoring is performed based on the number of faults, the number of power outages, the duration of power outages, the duration of power restoration, etc. to obtain a score for the reliability of power supply in the distribution network. Measuring the reliability of power supply by scoring requires a large amount of manpower to score and evaluate the reliability. On the other hand, manual scoring is greatly affected by subjective factors, making it difficult to comprehensively and accurately evaluate the reliability of power supply. Summary of the invention

[0004] The present invention provides a method and device for evaluating the reliability of power supply in a substation based on a neural network model, so as to solve the problem that manual scoring of power supply reliability based on collected data requires a large amount of manpower and has a large subjective influence, resulting in inaccurate scoring.

[0005] In a first aspect, the present invention provides a method for evaluating the reliability of power supply in a substation area based on a neural network model, comprising:

[0006] Upon receiving a reliability assessment request, determining a target substation in the distribution network;

[0007] Acquire historical power supply data, historical operation data of switchgear, topology line structure data and historical fault data of the target substation;

[0008] Inputting the historical power supply data, historical action data, topology line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation;

[0009] A power supply reliability level of the target substation is determined based on the reliability score.

[0010] Optionally, obtaining historical power supply data of the target substation, historical operation data of the switchgear, topology line structure data, and historical fault data includes:

[0011] Acquire the daily load curve, daily power supply voltage curve of the target substation and rated parameters of the distribution transformer of the target substation as historical power supply data;

[0012] Acquire the action time and action result of the switch device under the target substation as historical action data;

[0013] Obtaining a topological line map of the target area;

[0014] The fault type, fault time, fault elimination time, and power restoration time of various faults in the target substation are obtained as historical fault data.

[0015] Optionally, before inputting the historical power supply data, historical action data, topology line structure data and historical fault data into the reliability scoring model, the method further includes:

[0016] The historical power supply data, the historical action data of the switch device, the topological line structure data and the historical fault data are preprocessed to obtain the historical power supply data, the historical action data, the topological line structure data and the historical fault data in a standard format.

[0017] Optionally, the reliability scoring model includes a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network, the power supply feature extraction network is connected to the first scoring network, the switch action feature extraction network is connected to the second scoring network, the topology feature extraction network is connected to the third scoring network, the fault feature extraction network is connected to the fourth scoring network, the first scoring network, the second scoring network, the third scoring network and the fourth scoring network are all connected to the total scoring prediction network, the historical power supply data, the historical action data, the topology line structure data and the historical fault data are input into the reliability scoring model, and the power supply reliability score of the target substation is obtained, including:

[0018] Inputting the historical power supply data, historical action data, topology line structure data and historical fault data into the power supply feature extraction network, switch action feature extraction network, topology feature extraction network and fault feature extraction network respectively;

[0019] Extracting power supply features from the historical power consumption data in the power supply feature extraction network, extracting switch action features from the historical action data in the switch action feature extraction network, extracting topology features from the topology line structure data in the topology feature extraction network, and extracting fault features from the historical fault data in the fault feature extraction network;

[0020] In the first scoring network, the power supply characteristics are scored to obtain a power supply stability score, in the second scoring network, the switch action characteristics are scored to obtain a switch reliability score, in the third scoring network, the topology characteristics are scored to obtain a topology line reliability score, and in the fourth scoring network, the fault characteristics are scored to obtain a performance reliability score;

[0021] In the total score prediction network, the power supply reliability score of the target substation is predicted according to the power supply stability score, the switch reliability score, the topology line reliability score and the performance reliability score.

[0022] Optionally, the reliability scoring model is trained in the following manner:

[0023] Constructing a training data set, wherein a sample data in the training data set includes power supply data, switch device operation data, topology line structure data, and fault data of a sample substation on a daily basis, wherein the power supply data, switch device operation data, topology line structure data, and fault data in each sample data are respectively annotated with a first score, a second score, a third score, and a fourth score, and the sample substation is annotated with a fifth score of power supply reliability;

[0024] Constructing a reliability scoring model including a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network;

[0025] Randomly extract sample data from the training data set and input it into the reliability scoring model, respectively obtain a sixth score, a seventh score, an eighth score, and a ninth score from the first scoring network, the second scoring network, the third scoring network, and the fourth scoring network, and obtain a tenth score of the power supply reliability of the sample substation from the total score prediction network;

[0026] Calculating the loss rate using the first to the tenth scores;

[0027] Determine whether the training stop condition is met;

[0028] If so, determining that the reliability scoring model has completed training;

[0029] If not, the model parameters of the reliability scoring model are adjusted according to the loss rate, and the process returns to the step of randomly extracting sample data from the training data set and inputting it into the reliability scoring model.

[0030] Optionally, using the first score to the tenth score to calculate the loss rate includes:

[0031] Calculating a power supply stability score loss rate using the first score and the sixth score;

[0032] Calculating a switch reliability score loss rate using the second score and the seventh score;

[0033] Calculating the topology line reliability score loss rate using the third score and the eighth score;

[0034] Calculating a performance reliability score loss rate using the fourth score and the ninth score;

[0035] Calculating the total score loss rate using the fifth score and the tenth score;

[0036] According to the claim, adjusting the model parameters of the reliability scoring model according to the loss rate comprises:

[0037] Adjusting network parameters of the power supply feature extraction network and the first scoring network according to the power supply stability scoring loss rate;

[0038] Adjusting network parameters of the switch action feature extraction network and the second scoring network according to the switch reliability score loss rate;

[0039] Adjusting the network parameters of the topology feature extraction network and the third scoring network according to the topology line reliability scoring loss rate;

[0040] Adjusting network parameters of the fault feature extraction network and the fourth scoring network according to the performance reliability scoring loss rate;

[0041] The network parameters of the total score prediction network are adjusted according to the total loss rate.

[0042] Optionally, determining the power supply reliability level of the target substation based on the reliability score includes:

[0043] The level matching the reliability score is searched in a preset score-level comparison table as the power supply reliability level of the target substation.

[0044] In a second aspect, the present invention provides a device for evaluating the reliability of power supply in a substation area based on a neural network model, comprising:

[0045] A target substation determination module is used to determine a target substation in the distribution network when receiving a reliability assessment request;

[0046] A data acquisition module, used to acquire the historical power supply data of the target substation, the historical action data of the switchgear, the topological line structure data and the historical fault data;

[0047] A scoring module, used for inputting the historical power supply data, historical action data, topological line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation;

[0048] A reliability level determination module is used to determine the power supply reliability level of the target substation based on the reliability score.

[0049] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0050] at least one processor; and

[0051] a memory communicatively connected to the at least one processor; wherein,

[0052] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the substation power supply reliability assessment method based on the neural network model as described in any one of the first aspects of the present invention.

[0053] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a processor to implement the method for evaluating power supply reliability of a substation based on a neural network model as described in any one of the first aspect of the present invention when executed.

[0054] The embodiment of the present invention determines the target substation upon receiving a reliability assessment request, obtains historical power supply data, historical action data of the switch device, topological line structure data and historical fault data of the target substation, inputs the historical power supply data, historical action data, topological line structure data and historical fault data into a reliability scoring model, obtains the power supply reliability score of the target substation, and determines the power supply reliability level of the target substation based on the reliability score. This achieves that after the target substation is determined through a request operation, the historical power supply data, historical action data, topological line structure data and historical fault data of the target substation are obtained and input into the reliability scoring model to obtain the power supply reliability score of the target substation, so as to determine the corresponding reliability level. There is no need to manually collect data and then perform scoring, which saves manpower and eliminates the subjective influence of manual scoring. The power supply reliability score of the target substation can be accurately and objectively obtained, thereby accurately determining the reliability level, providing an accurate reference for system maintenance of the target substation.

[0055] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0057] Figure 1 It is a flow chart of a method for evaluating the reliability of power supply in a substation area based on a neural network model provided in the first embodiment of the present invention;

[0058] Figure 2 It is a flow chart of a method for evaluating the reliability of power supply in a substation area based on a neural network model provided in the second embodiment of the present invention;

[0059] Figure 3 is a schematic diagram of a network structure of a reliability scoring model in an embodiment of the present invention;

[0060] Figure 4 It is a structural schematic diagram of a power supply reliability assessment device for a substation based on a neural network model provided in Embodiment 3 of the present invention;

[0061] Figure 5 It is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0063] Embodiment 1

[0064] Figure 1 This is a flowchart of a method for evaluating the reliability of power supply in a distribution area based on a neural network model provided in the first embodiment of the present invention. This embodiment can be applied to the situation where the reliability of power supply in a distribution area is evaluated. The method can be performed by a device for evaluating the reliability of power supply in a distribution area based on a neural network model. The device for evaluating the reliability of power supply in a distribution area based on a neural network model can be implemented in the form of hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the power supply reliability assessment method based on the neural network model includes:

[0065] S101. When a reliability evaluation request is received, a target substation in a distribution network is determined.

[0066] In this embodiment, the reliability assessment can be an assessment of the power supply reliability of the substation. In one example, the reliability assessment request can be triggered by a preset trigger time. For example, when the number of fault work orders in the substation reaches a preset number within a preset time period (such as 1 day, 1 week, etc.), a reliability assessment request for the substation is triggered. For another example, when the number of complaints from users in the substation reaches a preset number, a reliability assessment request for the substation is triggered. It can also be possible to set a scheduled time at the same time to trigger the reliability request for the substation.

[0067] In another embodiment, an interactive interface may be provided in which the user may input or select a target area, select a time range for analysis, and trigger a reliability assessment request for the target area, thereby determining the ID of the target area and the time range for the reliability assessment from the reliability assessment request.

[0068] S102, obtaining historical power supply data of the target substation, historical operation data of the switchgear, topological line structure data and historical fault data.

[0069] In one embodiment, historical power supply data of the target substation, historical action data of the switch equipment, topological line structure data and historical fault data can be obtained in various systems through the ID of the target substation. Among them, the daily load curve of the target substation (which can be the load curve for each hour of 24 hours a day), the daily power supply voltage curve (which can be the output voltage curve of the substation distribution transformer for each hour of 24 hours a day) and the rated parameters of the distribution transformer of the target substation (rated load, voltage and other data of the substation distribution transformer) can be obtained as historical power supply data. Of course, the historical power supply data can also include current, flow and other data.

[0070] The action time and action results of the switching equipment under the target substation are obtained as historical action data, which can include the action time (such as tripping and closing duration) and action results (tripping and closing success or failure) of the switching equipment such as circuit breakers and relays under the target substation, and of course, the temperature of the switching equipment and other data.

[0071] Obtain the topological line map of the target substation, which is an electronic diagram of the transmission line under the target substation. The topological line map includes distribution equipment (such as distribution transformers, switchgear, current and voltage transformers, etc.) and line data connecting the distribution equipment (including wire diameter, withstand voltage and current parameters, etc.).

[0072] The fault types (short circuit fault, ground fault, overcurrent fault, overvoltage fault, etc.), fault time (the time when the fault occurs), fault elimination time (the time when the fault is eliminated), and power restoration time (the time from power outage due to fault maintenance to power restoration) of various faults in the target substation are obtained as historical fault data.

[0073] S103, inputting historical power supply data, historical operation data, topological line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation.

[0074] Specifically, the present embodiment can pre-train a reliability scoring model, which outputs a power supply reliability score after inputting data. The higher the power supply reliability score, the higher the power supply reliability of the target substation. In an example, the reliability scoring model can score historical power supply data, historical action data, topological line structure data, and historical fault data respectively to obtain scores in multiple dimensions, and then determine the power supply reliability score of the target substation by combining the scores in multiple dimensions. Scoring data from multiple dimensions can more comprehensively evaluate the power supply reliability of the target substation and improve the reliability and accuracy of the power supply reliability score.

[0075] S104. Determine the power supply reliability level of the target substation based on the reliability score.

[0076] In one embodiment, the power supply reliability assessment can be expressed in levels, for example, it can be divided into multiple levels, each level corresponds to a reliability score range, and the corresponding power supply reliability level can be determined by the reliability score of the target substation. System maintenance personnel can use the reliability level as a reference to specify inspection plans, maintenance plans, etc. For example, when the reliability level is low, the inspection intensity can be increased in the inspection plan.

[0077] The embodiment of the present invention determines the target substation upon receiving a reliability assessment request, obtains historical power supply data, historical action data of the switch device, topological line structure data and historical fault data of the target substation, inputs the historical power supply data, historical action data, topological line structure data and historical fault data into a reliability scoring model, obtains the power supply reliability score of the target substation, and determines the power supply reliability level of the target substation based on the reliability score. This achieves that after the target substation is determined through a request operation, the historical power supply data, historical action data, topological line structure data and historical fault data of the target substation are obtained and input into the reliability scoring model to obtain the power supply reliability score of the target substation, so as to determine the corresponding reliability level. There is no need to manually collect data and then perform scoring, which saves manpower and eliminates the subjective influence of manual scoring. The power supply reliability score of the target substation can be accurately and objectively obtained, thereby accurately determining the reliability level, providing an accurate reference for system maintenance of the target substation.

[0078] Embodiment 2

[0079] Figure 2 A flowchart of a method for evaluating the reliability of power supply in a substation area based on a neural network model is provided in the second embodiment of the present invention. The embodiment of the present invention is optimized on the basis of the above-mentioned first embodiment. Figure 2 As shown, the power supply reliability assessment method based on the neural network model includes:

[0080] S201. When a reliability evaluation request is received, a target substation in the distribution network is determined.

[0081] When receiving a reliability evaluation request, this embodiment can obtain the station area ID of the target station area from the reliability evaluation request.

[0082] S202: Acquire historical power supply data of the target substation, historical operation data of the switchgear, topological line structure data, and historical fault data.

[0083] Specifically, the historical power supply data of the target substation, the historical action data of the switchgear, the topological line structure data and the historical fault data can be obtained in each system of the distribution network through the ID of the target substation.

[0084] S203, pre-processing the historical power supply data, the historical action data of the switch device, the topological line structure data and the historical fault data to obtain the historical power supply data, the historical action data, the topological line structure data and the historical fault data in a standard format.

[0085] In this embodiment, the data preprocessing can be standardized and unified format processing. For example, the daily load curve and daily power supply voltage curve in the historical power supply data can be represented by load sequence and voltage sequence. For example, the preprocessed daily load curve can be Load (Load0, Load1, ..., Load24) to represent the load at each point in 24 hours a day. Similarly, the daily load curve can be V (V0, V1, ..., V24) to represent the voltage at each point in 24 hours a day. The preprocessed historical power supply data can be D1 (Load0, Load1, ..., Load24, V0, V1, ..., V24, C), where C can be the rated load, voltage and other data of the distribution transformer in the target substation.

[0086] The historical action data of the switching device can be expressed as D2 (type, time, result), where type represents the type of the switching device, such as a circuit breaker, relay, air switch, etc., time represents the action duration of the switching device (such as the reaction time for closing or tripping), and result represents the action result, 1 represents a successful action, and 0 represents a failed action.

[0087] For topological line structure data, the node level can be represented by a structure tree. Each node can include the parameters of the distribution equipment of the node (such as voltage, current, rated power, and if it is a switch device, it also includes a constant, etc.), and the connection between the nodes represents the wire diameter of the line between the two distribution devices.

[0088] Historical fault data can be expressed as D3 (type, time1, time2, time3, num), where type represents the fault type, time1, time2, time3 represent the fault time, fault elimination time, and power restoration time respectively, and num represents the number of times this type of fault occurs.

[0089] S204, inputting historical power supply data, historical action data, topology line structure data and historical fault data into a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network and a fault feature extraction network respectively.

[0090] like Figure 3 As shown, the reliability scoring model of this embodiment includes a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network. The power supply feature extraction network is connected to the first scoring network, the switch action feature extraction network is connected to the second scoring network, the topology feature extraction network is connected to the third scoring network, the fault feature extraction network is connected to the fourth scoring network, and the first scoring network, the second scoring network, the third scoring network, and the fourth scoring network are all connected to the total scoring prediction network.

[0091] The reliability scoring model of this embodiment can be trained by the following steps:

[0092] S1. Construct a training data set, in which a sample data in the training data set includes the daily power supply data of the sample substation, the operation data of the switch device, the topological line structure data and the fault data. The power supply data, the operation data of the switch device, the topological line structure data and the fault data in each sample data are respectively marked with a first score, a second score, a third score and a fourth score, and the sample substation is marked with a fifth score of power supply reliability.

[0093] Exemplarily, a sample data is {D1(x1, y1), D2(x2, y2), D3(x3, y3), D4(x4, y4), y5}, D1(x1, y1) represents power supply data x1 and a first score y1, D2(x2, y2) represents action data x2 and a second score y2, D3(x3, y3) represents topology line structure data x3 and a third score y3, and D4(x4, y4) represents fault data x4 and a fourth score y4.

[0094] S2. Construct a reliability scoring model including a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network.

[0095] Specifically, we can construct Figure 3 The reliability scoring model shown in the figure is constructed and the network parameters of each network are initialized.

[0096] S3. Randomly extract sample data from the training data set and input it into the reliability scoring model, and obtain the sixth score, seventh score, eighth score, ninth score from the first scoring network, second scoring network, third scoring network, and fourth scoring network respectively, and obtain the tenth score of the power supply reliability of the sample substation from the total scoring prediction network.

[0097] Specifically, the power supply data x1 is input into the power supply feature extraction network to obtain the power supply feature, which is then input into the first scoring network to obtain the sixth score y6, the action data x2 is input into the switch action feature extraction network to obtain the action feature, which is then input into the second scoring network to obtain the seventh score y7, the topology line structure data x3 is input into the topology feature extraction network to obtain the topology feature, which is then input into the third scoring network to obtain the eighth score y8, the fault data x4 is input into the fault feature extraction network to obtain the fault feature, which is then input into the fourth scoring network to obtain the ninth score y9, and then the sixth score y6, the seventh score y7, the eighth score y8, and the ninth score y9 are input into the total score prediction network to obtain the tenth score y10.

[0098] S4. Calculate the loss rate using the first to tenth scores.

[0099] Specifically, the first score y1 and the sixth score y6 are used to calculate the power supply stability score loss rate loss1, the second score y2 and the seventh score y7 are used to calculate the switch reliability score loss rate loss2, the third score y3 and the eighth score y8 are used to calculate the topology line reliability score loss rate loss3, the fourth score y4 and the ninth score y9 are used to calculate the performance reliability score loss rate loss4, and the fifth score y5 and the tenth score y10 are used to calculate the total score loss rate loss5, among which loss1, loss2, loss3, loss4, and loss5 can be calculated by the mean square error loss function.

[0100] S5. Determine whether the training stop condition is met.

[0101] In one embodiment, it can be determined whether the number of iterative training reaches a threshold number. If so, it is determined that the training stop condition is met and S6 is executed. If not, S7 is executed.

[0102] In another embodiment, the average of loss1, loss2, loss3, loss4, and loss5 may be calculated. If the average is less than a preset value, it is determined that the training stop condition is met and S6 is executed. If not, S7 is executed.

[0103] S6. Determine that the reliability scoring model has completed training.

[0104] When the stop training condition is met, it is determined that the accuracy of the reliability scoring model is sufficient, and the training of the reliability scoring model can be stopped.

[0105] S7. Adjust the model parameters of the reliability scoring model according to the loss rate.

[0106] Specifically, the network parameters of the power supply feature extraction network and the first scoring network can be adjusted according to the power supply stability scoring loss rate loss1, the network parameters of the switch action feature extraction network and the second scoring network can be adjusted according to the switch reliability scoring loss rate loss2, the network parameters of the topology feature extraction network and the third scoring network can be adjusted according to the topology line reliability scoring loss rate loss3, the network parameters of the fault feature extraction network and the fourth scoring network can be adjusted according to the performance reliability scoring loss rate loss4, and the network parameters of the total score prediction network can be adjusted according to the total loss rate loss5. The adjustment method can refer to the stochastic gradient descent, batch stochastic gradient descent, etc. in the prior art.

[0107] After completing the reliability scoring model training, the pre-processed historical power supply data, historical action data, topological line structure data and historical fault data of the target substation to be evaluated can be input into the power supply feature extraction network, switch action feature extraction network, topological feature extraction network and fault feature extraction network through the input layer.

[0108] S205. Extract power supply features from historical power consumption data in the power supply feature extraction network, extract switch action features from historical action data in the switch action feature extraction network, extract topology features from topology line structure data in the topology feature extraction network, and extract fault features from historical fault data in the fault feature extraction network.

[0109] Specifically, the power supply characteristics can be data such as daily maximum load, daily average load, daily minimum load, load rate, daily maximum voltage, daily minimum voltage, daily average voltage, voltage fluctuation rate, etc. The switch action characteristics can be data such as the maximum response time of the switch action and the action success rate, etc. The topological characteristics can be data such as the number of nodes, the wire diameter-load matching rate, and the rationality of the switch setting, etc. The fault characteristics can be data such as the fault type, the time period in which the fault occurred, the fault frequency, the maximum time for fault elimination, the average time, etc., the power outage duration, the power outage time period, the maximum time for power restoration, the average time for power restoration, etc., so that the power supply reliability of the substation can be measured through multi-dimensional characteristics.

[0110] S206. Score the power supply characteristics in the first scoring network to obtain a power supply stability score, score the switch action characteristics in the second scoring network to obtain a switch reliability score, score the topology characteristics in the third scoring network to obtain a topology line reliability score, and score the fault characteristics in the fourth scoring network to obtain a performance reliability score.

[0111] Specifically, the power supply stability score S1, the switch reliability score S2, the topology line reliability score S3 and the performance reliability score S4 can be recorded. Among them, the power supply stability score S1 reflects the power supply quality of the target substation, such as whether the power supply quality is stable from aspects such as load and power supply voltage. The switch reliability score S2 reflects the reliability of the target substation in responding to various events through switch actions during the power supply process from the length of switch action time and the success rate of switch action. The topology line reliability score S3 evaluates the selection of each node device in the target substation, whether the setting is reasonable, and whether the cable selection between the node devices is reasonable. The performance reliability score S4 evaluates the reliability of the target substation from the fault type, number of occurrences, occurrence time, fault elimination time, and power restoration time.

[0112] S207. Predict the power supply reliability score of the target substation according to the power supply stability score, the switch reliability score, the topology line reliability score, and the performance reliability score in the total score prediction network.

[0113] Specifically, the weights of the various scores may be set in the total score prediction network, and the weighted sum of the various scores may be calculated as the power supply reliability score of the target substation.

[0114] S208. Searching for a reliability score-grade matching grade in a preset score-grade comparison table as the power supply reliability grade of the target substation.

[0115] In one embodiment, the power supply reliability assessment can be expressed in levels, for example, it can be divided into multiple levels, each level corresponds to a reliability score range, and the corresponding power supply reliability level can be determined by the reliability score of the target substation. System maintenance personnel can use the reliability level as a reference to specify inspection plans, maintenance plans, etc. For example, when the reliability level is low, the inspection intensity can be increased in the inspection plan.

[0116] The embodiment of the present invention determines the target substation upon receiving a reliability assessment request, obtains historical power supply data, historical action data of the switch device, topological line structure data and historical fault data of the target substation, inputs the historical power supply data, historical action data, topological line structure data and historical fault data into a reliability scoring model, obtains the power supply reliability score of the target substation, and determines the power supply reliability level of the target substation based on the reliability score. This achieves that after the target substation is determined through a request operation, the historical power supply data, historical action data, topological line structure data and historical fault data of the target substation are obtained and input into the reliability scoring model to obtain the power supply reliability score of the target substation, so as to determine the corresponding reliability level. There is no need to manually collect data and then perform scoring, which saves manpower and eliminates the subjective influence of manual scoring. The power supply reliability score of the target substation can be accurately and objectively obtained, thereby accurately determining the reliability level, providing an accurate reference for system maintenance of the target substation.

[0117] Furthermore, after extracting power supply characteristics, switch action characteristics, topology characteristics and fault characteristics through multiple feature extraction networks for multi-dimensional scoring, the power supply reliability score of the target substation is determined by combining the multiple dimensional scores. This enables a comprehensive power supply reliability assessment of the target substation from the multiple dimensional characteristics of the target substation, and the assessment result is more accurate.

[0118] Embodiment 3

[0119] Figure 4 The schematic diagram of the structure of a power supply reliability assessment device based on a neural network model provided in the third embodiment of the present invention. Figure 4 As shown, the power supply reliability assessment device based on the neural network model includes:

[0120] The target substation determining module 401 is used to determine the target substation in the distribution network when receiving the reliability evaluation request;

[0121] A data acquisition module 402 is used to acquire historical power supply data, historical operation data of switchgear, topology line structure data and historical fault data of the target substation;

[0122] A scoring module 403 is used to input the historical power supply data, historical action data, topology line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation;

[0123] The reliability level determination module 404 is used to determine the power supply reliability level of the target substation based on the reliability score.

[0124] Optionally, the data acquisition module 402 includes:

[0125] A historical power supply data acquisition unit, used to acquire the daily load curve, daily power supply voltage curve of the target substation and the rated parameters of the distribution transformer of the target substation as historical power supply data;

[0126] A historical action data acquisition unit, used to acquire the action time and action result of the switchgear under the target substation as historical action data;

[0127] A topology line map acquisition unit, used to acquire a topology line map of the target area;

[0128] The historical fault data acquisition unit is used to acquire the fault type, fault time, fault elimination time, and power restoration time of various faults in the target substation as historical fault data.

[0129] Optionally, also include:

[0130] The data preprocessing module is used to preprocess the historical power supply data, the historical action data of the switch device, the topological line structure data and the historical fault data to obtain the historical power supply data, the historical action data, the topological line structure data and the historical fault data in a standard format.

[0131] Optionally, the reliability scoring model includes a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network, the power supply feature extraction network is connected to the first scoring network, the switch action feature extraction network is connected to the second scoring network, the topology feature extraction network is connected to the third scoring network, the fault feature extraction network is connected to the fourth scoring network, the first scoring network, the second scoring network, the third scoring network, and the fourth scoring network are all connected to the total scoring prediction network, and the scoring module 403 includes:

[0132] A data input unit, used to input the historical power supply data, historical action data, topology line structure data and historical fault data into the power supply feature extraction network, switch action feature extraction network, topology feature extraction network and fault feature extraction network respectively;

[0133] a feature extraction unit, configured to extract power supply features from the historical power consumption data in the power supply feature extraction network, extract switch action features from the historical action data in the switch action feature extraction network, extract topology features from the topology line structure data in the topology feature extraction network, and extract fault features from the historical fault data in the fault feature extraction network;

[0134] a scoring unit, configured to score the power supply characteristics in the first scoring network to obtain a power supply stability score, score the switch action characteristics in the second scoring network to obtain a switch reliability score, score the topology characteristics in the third scoring network to obtain a topology line reliability score, and score the fault characteristics in the fourth scoring network to obtain a performance reliability score;

[0135] A score prediction unit is used to predict the power supply reliability score of the target substation according to the power supply stability score, the switch reliability score, the topology line reliability score and the performance reliability score in the total score prediction network.

[0136] Optionally, the reliability scoring model is trained by the following modules:

[0137] A training data construction module is used to construct a training data set, wherein a sample data in the training data set includes power supply data, switch device action data, topology line structure data and fault data of a sample substation on a daily basis, and the power supply data, switch device action data, topology line structure data and fault data in each sample data are respectively annotated with a first score, a second score, a third score and a fourth score, and the sample substation is annotated with a fifth score of power supply reliability;

[0138] A model building module, used to build a reliability scoring model including a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network;

[0139] A multi-dimensional scoring module, used to randomly extract sample data from the training data set and input it into the reliability scoring model, and obtain a sixth score, a seventh score, an eighth score, and a ninth score from the first scoring network, the second scoring network, the third scoring network, and the fourth scoring network, respectively, and obtain a tenth score of the power supply reliability of the sample substation from the total score prediction network;

[0140] A loss rate calculation module, used to calculate the loss rate using the first score to the tenth score;

[0141] The training condition judgment module is used to judge whether the training stop condition is met; if so, the training completion determination module is executed; if not, the model parameter adjustment module is executed;

[0142] A training completion determination module is used to determine that the reliability scoring model has completed training;

[0143] If not, the model parameter adjustment module is used to adjust the model parameters of the reliability scoring model according to the loss rate and return to the multi-dimensional scoring module.

[0144] Optionally, the loss rate calculation module includes:

[0145] A first loss rate calculation unit, configured to calculate a power supply stability score loss rate using the first score and the sixth score;

[0146] A second loss rate calculation unit, configured to calculate a switch reliability score loss rate using the second score and the seventh score;

[0147] A third loss rate calculation unit, configured to calculate a topology line reliability score loss rate using the third score and the eighth score;

[0148] a fourth loss rate calculation unit, configured to calculate a performance reliability score loss rate using the fourth score and the ninth score;

[0149] A total loss rate calculation unit, used to calculate a total score loss rate using the fifth score and the tenth score;

[0150] The model parameter adjustment module includes:

[0151] A first adjustment unit, configured to adjust network parameters of the power supply feature extraction network and the first scoring network according to the power supply stability scoring loss rate;

[0152] A second adjustment unit, configured to adjust network parameters of the switch action feature extraction network and the second scoring network according to the switch reliability scoring loss rate;

[0153] A third adjustment unit, configured to adjust network parameters of the topology feature extraction network and the third scoring network according to the topology line reliability scoring loss rate;

[0154] A fourth adjustment unit, configured to adjust network parameters of the fault feature extraction network and the fourth scoring network according to the performance reliability scoring loss rate;

[0155] A fifth adjustment unit is used to adjust the network parameters of the total score prediction network according to the total loss rate.

[0156] Optionally, the reliability level determination module 404 includes:

[0157] The reliability grade searching unit is used to search for a grade matching the reliability score in a preset score-grade comparison table as the power supply reliability grade of the target substation.

[0158] The device for evaluating the reliability of power supply in a substation based on a neural network model provided in an embodiment of the present invention can execute the method for evaluating the reliability of power supply in a substation based on a neural network model provided in any embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method.

[0159] Embodiment 4

[0160] Figure 5 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0161] like Figure 5 As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0162] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0163] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 41 executes the various methods and processes described above, such as a power supply reliability assessment method for a substation based on a neural network model.

[0164] In some embodiments, the method for evaluating the reliability of power supply in a substation based on a neural network model may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the method for evaluating the reliability of power supply in a substation based on a neural network model described above may be performed. Alternatively, in other embodiments, the processor 41 may be configured to execute the method for evaluating the reliability of power supply in a substation based on a neural network model by any other appropriate means (e.g., by means of firmware).

[0165] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0166] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0167] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0168] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0169] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0170] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0171] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0172] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating the reliability of power supply in a substation area based on a neural network model, characterized in that: include: Upon receiving a reliability assessment request, determining a target substation in the distribution network; Acquire historical power supply data, historical operation data of switchgear, topology line structure data and historical fault data of the target substation; Inputting the historical power supply data, historical action data, topology line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation; A power supply reliability level of the target substation is determined based on the reliability score.

2. The method according to claim 1, characterized in that Obtaining historical power supply data, historical action data of switchgear, topology line structure data and historical fault data of the target substation, including: Acquire the daily load curve, daily power supply voltage curve of the target substation and rated parameters of the distribution transformer of the target substation as historical power supply data; Acquire the action time and action result of the switch device under the target substation as historical action data; Obtaining a topological line map of the target area; The fault type, fault time, fault elimination time, and power restoration time of various faults in the target substation are obtained as historical fault data.

3. The method according to claim 1, characterized in that Before inputting the historical power supply data, historical action data, topology line structure data and historical fault data into the reliability scoring model, the method further includes: The historical power supply data, the historical action data of the switch device, the topological line structure data and the historical fault data are preprocessed to obtain the historical power supply data, the historical action data, the topological line structure data and the historical fault data in a standard format.

4. The method according to any one of claims 1 to 3, characterized in that: The reliability scoring model includes a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network. The power supply feature extraction network is connected to the first scoring network, the switch action feature extraction network is connected to the second scoring network, the topology feature extraction network is connected to the third scoring network, the fault feature extraction network is connected to the fourth scoring network, the first scoring network, the second scoring network, the third scoring network and the fourth scoring network are all connected to the total scoring prediction network. The historical power supply data, historical action data, topology line structure data and historical fault data are input into the reliability scoring model to obtain the power supply reliability score of the target substation, including: Inputting the historical power supply data, historical action data, topology line structure data and historical fault data into the power supply feature extraction network, switch action feature extraction network, topology feature extraction network and fault feature extraction network respectively; Extracting power supply features from the historical power consumption data in the power supply feature extraction network, extracting switch action features from the historical action data in the switch action feature extraction network, extracting topology features from the topology line structure data in the topology feature extraction network, and extracting fault features from the historical fault data in the fault feature extraction network; In the first scoring network, the power supply characteristics are scored to obtain a power supply stability score, in the second scoring network, the switch action characteristics are scored to obtain a switch reliability score, in the third scoring network, the topology characteristics are scored to obtain a topology line reliability score, and in the fourth scoring network, the fault characteristics are scored to obtain a performance reliability score; In the total score prediction network, the power supply reliability score of the target substation is predicted according to the power supply stability score, the switch reliability score, the topology line reliability score and the performance reliability score.

5. The method according to claim 4, characterized in that The reliability scoring model is trained in the following way: Constructing a training data set, wherein a sample data in the training data set includes power supply data, switch device operation data, topology line structure data, and fault data of a sample substation on a daily basis, wherein the power supply data, switch device operation data, topology line structure data, and fault data in each sample data are respectively annotated with a first score, a second score, a third score, and a fourth score, and the sample substation is annotated with a fifth score of power supply reliability; Constructing a reliability scoring model including a power supply feature extraction network, a switch action feature extraction network, a topology feature extraction network, a fault feature extraction network, a first scoring network, a second scoring network, a third scoring network, a fourth scoring network and a total scoring prediction network; Randomly extract sample data from the training data set and input it into the reliability scoring model, respectively obtain a sixth score, a seventh score, an eighth score, and a ninth score from the first scoring network, the second scoring network, the third scoring network, and the fourth scoring network, and obtain a tenth score of the power supply reliability of the sample substation from the total score prediction network; Calculating the loss rate using the first to the tenth scores; Determine whether the training stop condition is met; If so, determining that the reliability scoring model has completed training; If not, the model parameters of the reliability scoring model are adjusted according to the loss rate, and the process returns to the step of randomly extracting sample data from the training data set and inputting it into the reliability scoring model.

6. The method according to claim 5, characterized in that The loss rate is calculated using the first score to the tenth score, including: Calculating a power supply stability score loss rate using the first score and the sixth score; Calculating a switch reliability score loss rate using the second score and the seventh score; Calculating the topology line reliability score loss rate using the third score and the eighth score; Calculating a performance reliability score loss rate using the fourth score and the ninth score; Calculating the total score loss rate using the fifth score and the tenth score; According to the claim, adjusting the model parameters of the reliability scoring model according to the loss rate comprises: Adjusting network parameters of the power supply feature extraction network and the first scoring network according to the power supply stability scoring loss rate; Adjusting network parameters of the switch action feature extraction network and the second scoring network according to the switch reliability score loss rate; Adjusting the network parameters of the topology feature extraction network and the third scoring network according to the topology line reliability scoring loss rate; Adjusting network parameters of the fault feature extraction network and the fourth scoring network according to the performance reliability scoring loss rate; The network parameters of the total score prediction network are adjusted according to the total loss rate.

7. The method according to any one of claims 1 to 3, characterized in that: Determining the power supply reliability level of the target substation based on the reliability score includes: The level matching the reliability score is searched in a preset score-level comparison table as the power supply reliability level of the target substation.

8. A device for evaluating the reliability of power supply in a substation area based on a neural network model, characterized in that: include: A target substation determination module is used to determine a target substation in the distribution network when receiving a reliability assessment request; A data acquisition module, used to acquire the historical power supply data of the target substation, the historical action data of the switchgear, the topological line structure data and the historical fault data; A scoring module, used for inputting the historical power supply data, historical action data, topological line structure data and historical fault data into a reliability scoring model to obtain a power supply reliability score for the target substation; A reliability level determination module is used to determine the power supply reliability level of the target substation based on the reliability score.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the substation power supply reliability assessment method based on the neural network model as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for evaluating the reliability of power supply in a substation based on a neural network model as described in any one of claims 1 to 7 when executed.

Images