An adaptive logic analysis method and system for power grid overload faults

Abstract

The application discloses a kind of power grid overload fault self-adapting logic analysis method and system, it is related to power grid operation fault analysis technical field, it includes the overload situation of target equipment based on multi-source data group identification and the reason type of overload and thermal stability state;At least one candidate disposal strategy is generated in combination with the reason type of overload of the target equipment and thermal stability state, carries out power flow calculation deduction in preset power grid range for each candidate disposal strategy, calculates the load rate of target equipment and all associated receiving equipment after load transfer, obtains new load rate result, obtains optimal strategy set based on new load rate result.It utilizes multi-source data fusion diagnosis overload reason and thermal stability state, and respectively carries out global power flow calculation deduction in the case where different strategies are generated, finally according to deduction result screening strategy, to overcome the blindness problem of relying on static fixed value judgment, disposal relies on artificial experience and lacks global risk prediction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid operation fault analysis, in particular to a self-adaptive logic analysis method and system for power grid overload fault. BACKGROUND

[0002] Overload refers to the current or power carried by a device exceeding its safety limit. If not handled in time, it may accelerate the insulation aging of the device, shorten its life, or even cause the device to burn out, trip, or even lead to a large-scale power outage. Currently, when the current or power of a device exceeds the limit, dispatchers take a series of preset control measures based on operating procedures and experience. However, this method has significant limitations, mainly lacking real-time, automatic, and accurate global deduction and evaluation of the risk of cascading failures caused by the disposal strategy during the process, which may lead to blind disposal and may cause systemic risks.

[0003] In the prior art, on the one hand, it mainly relies on whether the electrical quantity (such as current and power) exceeds the static limit to determine "overload", which cannot reflect the real operating state of the device. On the other hand, even if the cascading risk of the strategy is considered during temporary handling of device overload, the strategy is not accurately deduced and controlled, but only relies on experience to estimate and judge, which lacks accuracy and effectiveness. Especially, the reasons for device overload are various, and the weak links and evolution trends of the power grid corresponding to different reasons are completely different. Without a self-adaptive global impact evaluation mechanism, the cascading risk may become uncontrollable. SUMMARY

[0004] The purpose of the present application is to provide a self-adaptive logic analysis method and system for power grid overload fault, which diagnoses the overload reason and thermal stability state by using multi-source data fusion, performs global power flow calculation deduction under different strategies, and finally selects the strategy according to the deduction result, thereby overcoming the blindness of relying on static value judgment, disposal relying on artificial experience, and lack of global risk prediction.

[0005] Embodiments of the present application are implemented as follows:

[0006] In a first aspect, a self-adaptive logic analysis method for power grid overload fault is provided, comprising the following steps:

[0007] S100: Real-time acquisition of a multi-source data set of the target device, identification of the heavy overload condition of the target device and the cause type of the heavy overload based on the multi-source data set; wherein the multi-source data set includes electrical quantity data, physical quantity data, environmental quantity data, and power grid topology data; and the cause type includes natural load growth, fault flow transfer, and environmental change.

[0008] S200: Calculation of the real-time thermal stability margin of the target device based on the multi-source data set, determination of the remaining safety time of the target device based on the real-time thermal stability margin, and judgment of the thermal stability state of the target device according to the remaining safety time.

[0009] S300: Generation of at least one candidate treatment strategy in combination with the cause type and the thermal stability state of the heavy overload of the target device, calculation of the flow of the preset power grid range for each candidate treatment strategy, and prediction of the new load rate result of the preset power grid range after execution of the candidate treatment strategy; wherein when the flow of the preset power grid range is calculated for each candidate treatment strategy, the load rate of the target device and all associated receiving devices after load transfer is calculated, and the new load rate result of the preset power grid range is synthesized in combination with all load rates.

[0010] S400: Screening of all candidate treatment strategies based on the new load rate result to obtain an optimal strategy set.

[0011] In some optional embodiments, when the flow of the preset power grid range is calculated for each candidate treatment strategy, the load rate of the target device and all associated receiving devices after load transfer includes the following steps: construction of a network impedance model containing the target device and all associated receiving devices, execution of the corresponding candidate treatment strategy in the network impedance model to generate a deduced network topology structure; based on the deduced network topology structure, the load data of the current preset power grid range is calculated by flow calculation deduction to obtain the deduced power value of the target device and all associated receiving devices, and the load rate is calculated in combination with the respective deduced power value and dynamic allowable capacity of the target device and all associated receiving devices.

[0012] In some optional embodiments, the respective dynamic allowable capacity of the target device and all associated receiving devices includes the following steps: acquisition of real-time environmental quantity data of the target device and all associated receiving devices, calculation of the respective dynamic allowable capacity of the target device and all associated receiving devices in combination with a thermal stability model and real-time environmental quantity data; wherein the real-time environmental quantity data refers to environmental parameters respectively acquired by a plurality of monitoring points along the layout of the corresponding device, which are obtained by comprehensive calculation according to the environmental parameters of each monitoring point; and the environmental parameters include temperature, wind speed, and sunshine intensity.

[0013] In some alternative embodiments, the environment parameters respectively acquired at the plurality of monitoring points along the corresponding device layout include the following steps: determining all monitoring points covered in the upstream and downstream intervals of the corresponding device, screening out abnormal monitoring points among all monitoring points to obtain a set of valid monitoring points; selecting a collection mode based on the distribution form of the set of valid monitoring points to obtain respective environment parameters; wherein the collection mode includes collection objects, collection ranges, and collection proportions; wherein screening out abnormal monitoring points among all monitoring points includes the following steps: acquiring spatial continuity parameters of all monitoring points, calculating the validity confidence of each monitoring point according to the spatial continuity parameters of the monitoring points; acquiring parameter variation time series of all monitoring points, comparing and analyzing the variation trends of the parameter variation time series of the monitoring points to obtain variation trend abnormal values of each monitoring point; and determining whether to screen out in combination with the variation trend abnormal values and the validity confidence of each monitoring point.

[0014] In some alternative embodiments, the comparing and analyzing the variation trends of the parameter variation time series of the monitoring points to obtain variation trend abnormal values includes the following steps: determining a preset time window, based on the preset time window, the same length of all parameter variation time series is cut to obtain respective parameter comparison time series; similarity measurement is performed on all parameter comparison time series to obtain similarity measurement results, clustering analysis is performed based on the similarity measurement results to obtain variation trend characteristics; and difference comparison is performed between each group of parameter variation time series and the variation trend characteristics to obtain variation trend abnormal values.

[0015] In some alternative embodiments, the determination of the preset time window includes the following steps: obtaining abnormal change point positions in the parameter variation time series of each monitoring point, determining a basic time window value according to the corresponding abnormal change point positions; obtaining parameter variation rates in the parameter variation time series of each monitoring point, determining a window floating value according to the variation rates; obtaining individual time window values based on the basic time window value and the window floating value, and determining the preset time window according to the distribution of all individual time window values.

[0016] In some alternative embodiments, after the selection of the collection mode based on the distribution form of the set of valid monitoring points to obtain respective environment parameters, the following steps are further included: determining information of common monitoring points in each associated receiving device and the target device, increasing the value of the collection proportion of the monitoring points of the common monitoring point type in the collection mode; wherein the common monitoring point refers to a monitoring point used for data collection of at least two devices among the target device and all associated receiving devices.

[0017] In some alternative embodiments, the constructing the network impedance model comprising the target device and all associated receiving devices comprises the following steps: dividing the preset power grid range into different areas according to the electrical distance with the target device as the center, the different areas comprising a core area, a buffer area, and an external area; performing network impedance model construction for corresponding devices located in different areas: for the core area, constructing an accurate model comprising impedance coefficients of all devices in the area; for the buffer area, constructing a simplified model formed by merging secondary loads; for the external area, performing an external network equivalence step to construct an equivalent model; integrating the accurate model, the simplified model, and the equivalent model at the electrical boundary nodes of the preset power grid range to form the network impedance model.

[0018] In some alternative embodiments, the performing the external network equivalence step to construct the equivalent model for the external area comprises the following steps: determining the electrical boundary nodes between the buffer area and the external area, calculating the equivalent injection power at the boundary nodes, which is the boundary tie-line power result; forming the equivalent model of the external area by calculating the equivalent impedance from the boundary nodes to the external system; wherein the boundary tie-line power result and the derived power value together constitute a safety evaluation data set for candidate treatment strategy screening.

[0019] In a second aspect, an adaptive logic analysis system for power grid overload faults comprises:

[0020] A first analysis unit is configured to collect a multi-source data set of a target device in real time, identify an overload condition of the target device and a cause type of the overload based on the multi-source data set; wherein the multi-source data set comprises electrical quantity data, physical quantity data, environmental quantity data, and power grid topology data; and the cause type comprises natural load growth, fault power flow transfer, and environmental change;

[0021] A second analysis unit is configured to calculate a real-time thermal stability margin of the target device based on the multi-source data set, determine a remaining safety time of the target device based on the real-time thermal stability margin, and judge a thermal stability state of the target device according to the remaining safety time;

[0022] A first processing unit is configured to generate at least one candidate treatment strategy in combination with the cause type and the thermal stability state of the target device overload, perform power flow calculation derivation for a preset power grid range for each candidate treatment strategy, and predict a new load rate result of the preset power grid range after executing the candidate treatment strategy; wherein when performing power flow calculation derivation for the preset power grid range for each candidate treatment strategy, the load rate of the target device and all associated receiving devices after load transfer is calculated, and the new load rate result of the preset power grid range is synthesized in combination with all load rates;

[0023] A second processing unit is configured to filter all candidate processing strategies based on the new load rate result to obtain an optimal strategy set.

[0024] The embodiment of the present application has the following advantages:

[0025] The adaptive logic analysis method and system for power grid overload fault provided by the embodiment of the present application can pre-judge the overload by using real-time multi-source data of the target device, analyze the reason type of the overload, judge the real-time thermal stability margin by using real-time multi-source data to calculate the remaining safety time of the target device, and then obtain the thermal stability state of the target device. Then, the candidate disposal strategies are selected comprehensively according to the reason type of the overload and the thermal stability state, and the power flow calculation deduction is performed for each candidate disposal strategy. The load states of all devices involved after the change of the power flow are controlled globally, the problem of causing other cascading faults by solving a fault is avoided due to the single perspective, the load rate prediction of the target device and the associated receiving device after the change of the power flow is combined, the load states of all devices are further controlled, the pseudo-realism and reliability of the calculation in the deduction process are increased, and the high systematic risk caused by the traditional single strategy or the failure to control globally is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0027] Figure 1 The flowchart of the main steps of the analysis method provided by the embodiment of the present application;

[0028] Figure 2 The flowchart of one of the main steps S300 shown in the figure; Figure 1 The flowchart of the sub-step S310 of the step S300 shown in the figure;

[0029] Figure 3 The flowchart of the sub-step S310 of the step S300 shown in the figure; Figure 2 The flowchart of the sub-step S310 of the step S300 shown in the figure;

[0030] Figure 4 The block diagram of the analysis system provided by the embodiment of the present application.

[0031] Figure legend: 500-analysis system; 510-first analysis unit; 520-second analysis unit; 530-first processing unit; 540-second processing unit. DETAILED DESCRIPTION

[0032] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0033] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0034] It should be understood that the "system", "device" and / or "module" used in the present application is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0035] As shown in the present application and claims, unless the context clearly indicates otherwise, the words "one", "a", "an" and / or "the" do not refer to the singular, but can also include the plural. Generally, the terms "comprise" and "include" only indicate the inclusion of the steps and elements explicitly identified, and these steps and elements do not constitute an exclusive list, and the method or device can also include other steps or elements.

[0036] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or subsequent operations are not necessarily performed in sequence. On the contrary, each step can be processed in reverse order or simultaneously. Meanwhile, other operations can be added to these processes, or one or more steps of operations can be removed from these processes.

[0037] Embodiment: For the aspect of power grid overload fault, the "overload" is determined by whether the electrical quantity (such as current, power) exceeds the static limit value. This mode has some problems: for example, in the case of low ambient temperature and high wind speed, the actual current-carrying capacity (dynamic capacity) of the line can be much higher than the static limit; on the contrary, in the high temperature and windless environment, even if the current does not exceed the static limit, the equipment may face the risk of thermal stability damage due to poor heat dissipation, and the real-time thermal stability margin of the corresponding equipment cannot be calculated by combining real-time environmental parameters and accurate thermal circuit model to determine the severity of overload or how long it can be carried. Therefore, it is necessary to further analyze by real-time evaluation of the overload situation to obtain a more comprehensive overload degree.

[0038] On this basis, we adopt a real-time judgment of heavy overload and its degree, but even if the heavy overload degree data is obtained, only for this aspect to develop coping strategies, not to mention whether the strategy is targeted (especially the disposal strategy is out of touch with the fault reason, lack of pertinence, this predetermined or universal strategy is not deeply bound with the specific overload reason type); And it is easy to ignore the overall correlation effect, for example, in the strategy to alleviate the overload of A device, it may transfer a large amount of flow to B device, causing B device to be overloaded or approaching the stability limit. If the initial strategy cannot be dynamically adjusted or rejected according to the deduced "correlation influence degree" (especially the overload influence on other related devices), it is also impossible to evaluate whether the overall security level of the whole network is improved or decreased after the strategy is executed. Therefore, the embodiment provides an adaptive logic analysis method for heavy overload fault of power grid, to solve the problem that the heavy overload fault adopts a rough judgment standard, executes a fixed strategy with weak relevance to the root cause, and cannot accurately judge the chain effect of the strategy, so that the heavy overload disposal process of the power grid has a huge systematic risk.

[0039] For details, please refer to Figure 1 The adaptive logic analysis method for heavy overload fault of power grid provided by the embodiment comprises the following steps:

[0040] S100: Real-time acquisition of a multi-source data set of a target device, identification of heavy overload of the target device and a reason type of the heavy overload based on the multi-source data set; This step represents real-time data acquisition for the target device being monitored or patrolled, to obtain its related multi-source data set, wherein the multi-source data set at least includes electrical quantity data, physical quantity data, environmental quantity data and power grid topology data, to determine whether heavy overload occurs and the reason type of the heavy overload through these types of multi-source data, and the reason type includes natural load growth, fault flow transfer and environmental change; Specifically, for example, by comparing electrical quantity data (such as current, power) with the safety limit of the device, it is directly determined whether heavy overload occurs; Then cross-analyze other data sources to diagnose the overload root cause, if the power grid topology data shows that the adjacent device is tripped and the electrical quantity data increases sharply, it is determined as "fault flow transfer"; If the environmental quantity data (such as environmental temperature, solar intensity) is significantly abnormal and causes the physical quantity data (such as device temperature) to rise synchronously, it is determined as "environmental change"; If the electrical quantity and physical quantity data show a steady and slow growth trend and there is no abnormality, it is determined as "natural load growth", etc. The reason type of heavy overload is obtained through the above multi-source data set.

[0041] S200: calculating a real-time thermal stability margin of the target device based on the multi-source data set, determining a residual safety time of the target device based on the real-time thermal stability margin, and judging a thermal stability state of the target device according to the residual safety time; this step means that the real-time thermal stability margin of the target device is first calculated by means of the multi-source data set, for example, the electrical quantity data (such as load current) in the multi-source data set is taken as the main heat source, the physical quantity data (such as the current conductor temperature) is taken as the state feedback, and the environmental quantity data (such as the environmental temperature and wind speed) is taken as the heat dissipation boundary condition, which is substituted into the thermodynamic model describing the heat balance relationship of the device to calculate the real-time thermal stability margin (i.e. the difference between the current temperature of the device and the maximum allowable temperature) of the device under the current operating condition. The thermodynamic model can be constructed based on the heat equivalent principle, and the heat generation and heat dissipation process of the device is described as a dynamic balance system, the heat generation power of which is calculated from the real-time load current in the electrical quantity data and the device resistance, and the heat dissipation power is determined by the comprehensive heat dissipation coefficient (which can be determined by fitting the current temperature of the device in the physical quantity data, the environmental temperature and wind speed in the environmental quantity data, and the inherent heat capacity parameters of the device itself)

[0042] Then, the current load and environmental conditions are taken as inputs, the time required for the device temperature to rise to the limit value is predicted by solving the thermodynamic model, thereby determining the residual safety time, and finally comparing the residual safety time with the preset time threshold, thereby quantitatively judging that the device is in different thermal stability states such as safe, warning or emergency.

[0043] S300: generating at least one candidate treatment strategy in combination with the cause type and thermal stability state of the overload of the target device, performing power flow calculation deduction in a preset power grid range for each candidate treatment strategy, and predicting the new load rate result of the preset power grid range after executing the candidate treatment strategy; this step means that the cause type and thermal stability state of the overload obtained by the foregoing steps are matched to generate different coping strategies, that is, at least one candidate treatment strategy is generated, for example, if the diagnosis cause is fault power transfer and the thermal stability state is emergency, the generated candidate treatment strategy preferentially includes emergency load shedding or rapid switching of local distributed power supply; for example, if the cause is natural load growth and the state is warning, the generated candidate treatment strategy focuses on adjusting the power grid operation mode or starting the standby transformer; if the cause is environmental change, the strategies of enabling auxiliary cooling system or adjusting the power flow of adjacent lines to reduce the current of the target device are preferentially generated, so as to ensure that the generated strategy set has pertinence and timeliness.

[0044] Then each selected candidate treatment strategy is subjected to preset power grid range power flow calculation deduction, that is, power flow calculation deduction is performed in the predetermined power grid equipment area range (all electrical associated equipment of the target equipment is determined in advance according to the power grid topology data, so as to delimit the preset power grid range which needs to be deduced), which means that each candidate treatment strategy is quantified as the corresponding control parameter change in the power grid simulation model, the real-time collected multi-source data set is used as the initial operation mode, high-precision power flow calculation is performed in the preset power grid range, the new load rate result of the preset power grid range after executing the candidate treatment strategy is predicted, that is, by solving the power grid steady-state equation, the power flow distribution of all equipment (target equipment and each associated receiving equipment) in the preset power grid range after executing the candidate treatment strategy is accurately predicted, and the new load rate of each equipment is calculated to form the new load rate result representing the global influence after the strategy is executed.

[0045] On the basis of the above technical solutions, since the power flow change will involve the power distribution change of all associated equipment, when performing the power flow calculation deduction of the preset power grid range, the load transient response characteristics and load continuous change process of each equipment need to be specially considered, and the load of the target equipment is not simply transferred to the associated receiving equipment, but the candidate treatment strategy is quantified as the specific change of generator output, load switching or network topology in the power grid model, the power network differential algebraic equation containing all associated equipment is established, and then the equation is solved to simulate the power flow distribution change in the new steady-state process after the strategy is executed. Therefore, the power load of each equipment after the power flow change needs to be considered, so as to recalculate the new load rate of all equipment, that is, when performing the power flow calculation deduction of the preset power grid range for each candidate treatment strategy, the load rate of the target equipment and all associated receiving equipment after load transfer is calculated, and the new load rate result of the preset power grid range is synthesized by combining all load rates (discrete load rate data is weighted and fused as a whole data set to construct a comprehensive index which can comprehensively and quantitatively represent the safe operation level of the entire preset power grid range after executing the candidate treatment strategy, that is, the new load rate result)

[0046] S400: screening all candidate treatment strategies based on the new load rate result to obtain an optimal strategy set; this step means that the candidate treatment strategies corresponding to the new load rate result which can make all equipment in the preset power grid range operate more stably are extracted by the obtained new load rate result to obtain an optimal strategy set, and the optimal strategy set is used as the basis or one of the subsequent strategy implementation.

[0047] By the technical solution, the multi-source data such as electrical quantity, physical quantity, environmental quantity and power grid topology are fused, so that not only the overload phenomenon can be accurately identified, but also the root causes such as natural growth of load, fault flow transfer and environmental change can be synchronously diagnosed, thereby providing a targeted basis for subsequent strategy generation. In addition, by introducing a thermodynamic model based on the heat path equivalent principle, static load rate monitoring is converted into dynamic evaluation of real-time thermal stability margin and remaining safety time, the cause type is matched with the thermal stability state, and through fine power flow calculation deduction of each candidate strategy in a preset power grid range, global power flow distribution change after strategy execution is accurately simulated, and a new load rate result (comprehensive load rate index) reflecting the global safety level is synthesized. This mode of closed-loop processing chain of multi-source data diagnosis-thermal stability dynamic evaluation-strategy power flow deduction-global optimization screening can solve the chain risk caused by single strategy execution in the traditional mode relying on artificial experience, and also increases the degree of global control of strategy execution, thereby outputting an optimal strategy set that can better guarantee global adaptive operation, safe operation and stable operation.

[0048] On the basis of the above technical solution, the load rate of the target device and all associated receiving devices is generally calculated by the new power value and the rated capacity of each device. However, considering that the rated capacity is a parameter configured at the factory, the actual capacity will change according to different factors such as environmental changes, aging and reaction time. If only the rated capacity is considered for calculation, a more realistic load rate value cannot be obtained. Therefore, in the present embodiment, referring to Figure 2 , when performing power flow calculation deduction of each candidate treatment strategy in a preset power grid range, the load rate of the target device and all associated receiving devices after load transfer includes the following steps:

[0049] S310: Construct a network impedance model containing the target device and all associated receiving devices, execute the corresponding candidate treatment strategy in the network impedance model, and generate a deduction network topology structure. This step represents that a network impedance model of the corresponding device is first constructed (for example, the electrical connection relationship and device impedance parameters of the target device and all associated receiving devices are extracted based on the power grid topology data, and an electrical network impedance model expressed in the form of a node admittance matrix is established), and then the candidate treatment strategy is quantified as a corresponding change in switch state, load switching or adjustment of power supply output in the network impedance model, the connection relationship and injected current source term of the model are dynamically modified, thereby generating a deduction network topology structure reflecting the new operating mode after strategy execution, so as to provide a relatively real power grid structure basis for subsequent power flow calculation.

[0050] S320: Based on the deduction network topology, the load data of the current preset power grid range is carried out power flow calculation deduction, and the deduction power value of the target device and all associated receiving devices is obtained; this step means that the real-time collected load data is input as a boundary condition into the above-mentioned obtained deduction network topology, and the power distribution of the target device and all associated receiving devices under the new network structure after the execution of the candidate treatment strategy is calculated by solving the power grid power flow equation, so as to obtain the deduction power value of each device.

[0051] S330: Calculate the load rate in combination with the deduction power value of the target device and all associated receiving devices and the dynamic allowable capacity; this step means that the obtained deduction power value of each device is matched and calculated with the corresponding dynamic allowable capacity, which respectively obtains the load rate value of the target device and all associated receiving devices after the execution of the strategy by dividing the deduction power value by the dynamic allowable capacity and multiplying by the percentage. Among them, the dynamic allowable capacity is the instantaneous maximum carrying capacity obtained by correcting the rated capacity based on real-time environmental parameters and device running state. Compared with the rated capacity, the dynamic allowable capacity can dynamically correct the actual maximum carrying capacity of the device by integrating real-time environmental quantity data, physical quantity data and device aging state and other multi-source running parameters, so as to more truly reflect the instantaneous carrying limit of the device under the current running condition.

[0052] Through the above technical scheme, the network impedance model is constructed to generate the deduction network topology, the device deduction power value is obtained by combining the power flow calculation, and the load rate calculation is carried out by replacing the traditional rated capacity with the dynamic allowable capacity, so that the load rate evaluation result can reflect the dynamic influence of the actual running state such as environmental condition and device aging degree on the carrying capacity of the device in real time, which can significantly improve the authenticity and accuracy of the load rate calculation.

[0053] Considering that the dynamic allowable capacity is greatly affected by real-time environmental parameter changes, when determining the real-time environmental parameters, real-time environmental quantity data needs to be combined as the main basis for calculation. Specifically, obtaining the dynamic allowable capacity of the target device and all associated receiving devices includes the following steps:

[0054] S331: Collect real-time environmental quantity data of the target device and all associated receiving devices; this step means that the real-time environmental quantity data of the target device and all associated receiving devices in the analysis period is collected first. The real-time environmental quantity data is the environmental parameter obtained by the respective monitoring points along the layout of the corresponding device (the target device or the associated receiving device). The environmental parameter at least includes temperature, wind speed and sunshine intensity. The real-time environmental quantity data is obtained by comprehensive calculation according to the environmental parameter of each monitoring point.

[0055] S332: Calculate the dynamic allowable capacity of the target device and all associated receiving devices respectively by combining the thermal stability model and real-time environmental quantity data; this step represents using the thermal stability model (for example, based on the principle of thermal equilibrium, regarding the corresponding device as a thermodynamic system, and realizing it by constructing a differential equation describing the dynamic balance relationship between the heat generation and heat dissipation of the device) to substitute the real-time environmental quantity data to calculate the dynamic allowable capacity of the corresponding device, that is, inputting the real-time environmental quantity data as boundary conditions into the thermal stability model matched with the corresponding device, calculating the maximum carrying current that the device can bear under the thermal stability constraint condition of not exceeding the maximum allowable temperature, and then converting the current value into power to calculate the dynamic allowable capacity of the device under the current actual environmental conditions.

[0056] Through the above technical solution, the temperature, wind speed, and solar intensity and other environmental parameters are collected by multiple monitoring points and comprehensively calculated to obtain real-time environmental quantity data that can truly reflect the local microenvironment of the device, which is then input as boundary conditions into the thermal stability model constructed based on the principle of thermal equilibrium, the maximum carrying current under the thermal stability constraint condition is solved and converted into power to calculate the dynamic allowable capacity of the device, so that the dynamic allowable capacity evaluation result can respond to environmental changes in real time, and the accuracy and reliability of the load rate calculation are improved.

[0057] On the basis of the above technical solution, real-time environmental quantity data is related to the accuracy of the final dynamic allowable capacity result calculation, so the environmental parameters of the monitoring points need to be considered in a more adaptive mode, such as extracting more relevant monitoring point data for effective monitoring point data. Specifically, the environmental parameters obtained by the multiple monitoring points along the layout of the corresponding device include the following steps:

[0058] S3311: Determine all monitoring points covered in the upstream and downstream intervals of the corresponding device, exclude abnormal monitoring points in all monitoring points, and obtain an effective monitoring point set; this step represents first determining the monitoring point coverage range of the corresponding device (target device or associated receiving device) for which the dynamic allowable capacity is to be calculated, which mainly refers to the range of monitoring point layout that can collect data in the upstream and downstream electrical monitoring areas of the device. Then, the abnormal monitoring points among the monitoring points are excluded, and the remaining monitoring points constitute an effective monitoring point set. Among them, the abnormal monitoring points among all monitoring points are excluded, including the following steps:

[0059] obtain the spatial continuity parameters of all monitoring points, and calculate the validity confidence of each monitoring point according to the spatial continuity parameters; this step represents the continuity analysis on the spatial distribution of all monitoring points covered by the device to be analyzed, and obtains the spatial continuity parameters representing the "smoothness" and "physical rationality" of the "environmental parameter field" formed by all monitoring points in the spatial distribution, for example, a continuous environmental parameter field covering the area covered by the device can be generated by using a spatial interpolation algorithm, a spatial correlation index is constructed by calculating the standard residual error of the measured value of each monitoring point and the predicted value of the interpolation, and combining the spatial distance relationship between the point and the adjacent monitoring points, and finally the residual error statistics and the spatial correlation index are fused into the spatial continuity parameters representing the reliability of the data of the monitoring point, for example, three monitoring points A, B and C are evenly distributed at equal intervals on a straight line, and the temperatures measured by them are 25℃, 26℃ and 40℃, respectively. From the perspective of spatial continuity, the temperature of points A and B is gradually changed, and has good continuity, but the temperature of points B and C has a sharp jump, and the spatial continuity of point C is not high, and the relative distribution is more discrete. Finally, the validity confidence of each monitoring point in data extraction is selected according to the discrete value of the spatial continuity parameter of each monitoring point, and the higher the discrete value, the lower the validity confidence.

[0060] Then obtain the parameter change time sequence of all monitoring points, compare the change trend of each monitoring point, and obtain the change trend abnormal value of each monitoring point; this step represents that the historical environmental data of each monitoring point in the continuous time period is collected to form a parameter change time sequence (a parameter sequence arranged in time), and then the slope, fluctuation characteristics and fitting trend of the parameter change curve of each monitoring point in the preset time period are compared, the deviation degree of the data of each monitoring point from the average change of the whole region is calculated, and the deviation degree is quantized as the change trend abnormal value. Finally, whether to be screened out is judged according to the change trend abnormal value and the validity confidence of each monitoring point, that is, the change trend abnormal value and the validity confidence calculated by each monitoring point are used as the screening basis, for example, the change trend abnormal value and the validity confidence are high, and then the change trend abnormal value and the validity confidence are low, and so on, and the monitoring points meeting the standard can be reserved according to the needs.

[0061] S3312: selecting a collection mode based on the distribution form of the set of valid monitoring points, to obtain respective environmental parameters; wherein the collection mode includes a collection object, a collection range, and a collection proportion; this step represents that all remaining valid monitoring points are analyzed in a distribution form, and the collection strategy is dynamically adjusted according to the spatial distribution characteristics of the set of valid monitoring points; the collection modes of the concentrated or dispersed valid monitoring points are different, for example, when the set of valid monitoring points is concentratedly distributed (forming a high-density area), a high-density collection mode is adopted, the collection object is focused on the monitoring points in the core area, the collection range is reduced, and the parameter collection proportion (the proportion of collection of all monitoring points) of each point is increased; when the valid monitoring points are dispersedly distributed (there are more low-density areas), a wide-area collection mode is adopted, the collection range is expanded to the coverage area of all valid monitoring points, the collection object is determined according to the uniform distribution principle, and each point maintains the baseline collection proportion.

[0062] Through the above technical solution, the abnormal monitoring points caused by equipment failure or local micro-environmental interference can be effectively screened out by using the spatial continuity analysis and parameter change time sequence comparison double-checking mechanism, to ensure the physical rationality and reliability of the data source; at the same time, the collection mode is adaptively selected according to the distribution form (spatial distribution characteristics) of the set of valid monitoring points, a high-density collection is used in the concentrated area to capture the detailed features, and a wide-area collection is used in the dispersed area to ensure the spatial representation integrity, so that the pertinence and representativeness of environmental parameter collection are significantly improved, to lay a reliable data foundation for the accurate calculation of the subsequent dynamic allowable capacity.

[0063] In this embodiment, since the change trend abnormal value analysis needs to compare the parameter change time sequences of all monitoring points, the alignment accuracy needs to be considered to ensure the reliability of the comparison and analysis result. Specifically, the change trend comparison and analysis of the parameter change time sequences of the monitoring points, to obtain the change trend abnormal value of each monitoring point, includes the following steps:

[0064] A preset time window is determined, which can be determined in advance according to the data collection conditions of all monitoring points. Then, the parameter change time series of all parameters are cut off at the same length based on the preset time window, to obtain respective parameter comparison time series. At this time, all the cut-off parameter comparison time series can be aligned. Then, similarity measurement is performed on all the aligned parameter comparison time series, for example, dynamic time warping algorithm or Pearson correlation coefficient method is used to perform similarity measurement on the aligned parameter comparison time series, to obtain similarity measurement results representing the consistency of change rules between monitoring points. Finally, clustering analysis is performed based on the similarity measurement results to obtain the change trend feature, that is, the monitoring points with common change mode are merged into the same feature group through clustering algorithm, so as to extract the change trend feature reflecting the evolution rule of regional environmental parameters. The parameter change time series of each group is compared with the change trend feature in turn to obtain the change trend abnormal value, that is, the parameter change time series of each monitoring point is compared with the representative change trend feature obtained by clustering point by point, for example, residual sum of squares or dynamic time warping distance algorithm is used to quantify the deviation degree, and the deviation degree is normalized to obtain the change trend abnormal value reflecting the consistency of the monitoring point data and the regional overall change mode.

[0065] Through the above technical solution, the preset time window is used to realize the accurate alignment of the parameter change time series of multiple monitoring points, the dynamic time warping and clustering analysis are combined to extract the change trend feature of the regional environmental parameters, and the deviation degree of the parameter change time series of each monitoring point from the change trend feature is quantified, so that the abnormal data caused by equipment failure or local microenvironment can be effectively identified, and the quality and reliability of the environmental parameter collection are improved.

[0066] On the basis of the above technical solution, the preset time window is determined according to the data collection conditions of all monitoring points. Since the accuracy of abnormality judgment needs to be considered, the time point of each monitoring point abnormality and the degree of abnormality need to be identified, so that a window value size that can consider overall abnormality detection is selected. Specifically, the determination of the preset time window includes the following steps:

[0067] The abnormal change point positions in the parameter change time series of each monitoring point are obtained, and the basic time window value is determined according to the corresponding abnormal change point positions; that is, for the parameter change time series of each monitoring point, the point position of abnormal change is identified (for example, a sliding window algorithm or a threshold-based mutation detection method is used to scan the parameter change time series of each monitoring point in real time, and when the parameter value of the point position deviates from the historical mean value in the sliding window by more than a preset standard deviation multiple or a mutation threshold, the time point is marked as an abnormal change point position), and the basic time window value (a window covering the span of these abnormal points) is determined according to the time span of these abnormal change points.

[0068] The parameter change rate in the parameter change time sequence of each monitoring point is obtained, and a window floating value is determined according to the change rate; that is, a window floating value is determined according to the parameter change rate (jitter condition) of all parameters in the parameter change time sequence of the monitoring point, which is used to determine whether the window needs to be shortened or expanded. For example, the change amplitude between consecutive data points in the parameter change time sequence is calculated by using a linear fitting slope method to quantify the fluctuation intensity of the parameter change rate. When the standard deviation of the change rate exceeds the high fluctuation threshold, a negative window floating value is set to reduce the analysis window to capture the rapid change characteristics. When the standard deviation of the change rate is lower than the low fluctuation threshold, a positive window floating value is set to expand the analysis window to smooth random fluctuations.

[0069] The individual time window value is obtained based on the basic time window value and the window floating value, and the preset time window is determined according to the distribution of all individual time window values; that is, the individual time window value is calculated by using the basic time window value and the window floating value obtained for each monitoring point, which is used to represent the window preferred value of the monitoring point. Then, the preset time window is determined according to the distribution of all individual time window values. For example, statistical analysis is performed on the data set of all individual time window values to determine the core distribution interval, and the median of the interval is taken as the reference value. Then, the reference value is combined with the power grid dispatching period for integer optimization, so as to determine the preset time window that can take into account the characteristics of most monitoring points and the actual operation demand.

[0070] Through the above technical solution, the abnormal change point position in the parameter change time sequence of each monitoring point is identified to determine the basic time window value, the window floating value is dynamically adjusted according to the parameter change rate, and finally the preset time window that can take into account the comprehensiveness and accuracy of abnormal detection is determined based on the distribution characteristics of all individual time window values of the monitoring points, so as to improve the adaptive compatibility of the time window.

[0071] In the above technical solution, the collection strategy is mainly dynamically adjusted according to the spatial distribution characteristics of the effective monitoring point set. In this case, the environmental parameters monitored by an effective monitoring point may be suitable for environmental perception monitoring of multiple devices, so that they can all participate in the dynamic allowed capacity calculation of these devices. Therefore, the importance of collecting parameters needs to be considered in the case of multiple device perception monitoring of one monitoring point, especially the proportion of the collected parameters in all monitoring points. Specifically, after the environmental parameters of each device are obtained based on the distribution form of the effective monitoring point set, the following steps are further included:

[0072] Determine the information of the common monitoring point of the target device and each associated receiving device, which refers to the monitoring point used for data collection of at least two devices in the target device and all associated receiving devices. Then, the monitoring point determined as the common monitoring point type is increased in the collection proportion value in the collection mode, that is, the effective monitoring point belonging to the common monitoring point needs to increase the overall proportion of data collection (whether in the concentrated area or the dispersed area, the increase degree is selected according to the demand). Through the foregoing technical solution, the common monitoring point serving multiple devices is identified, and the collection proportion in the data collection mode is actively increased, which can effectively strengthen the data support role of the key shared node.

[0073] In constructing the network impedance model containing the target device and all associated receiving devices, it is considered that the associated devices with close electrical distance to the target device have more significant influence on the result in the power flow calculation, and a higher precision model needs to be used for representation; and with the increase of the electrical distance, the influence of the associated device on the target device will gradually weaken due to transmission attenuation and other factors. In order to ensure the calculation accuracy and improve the deduction efficiency, the embodiment adopts a differentiated modeling precision strategy for the target device and the associated receiving devices with different electrical distances, and finally integrates and constructs a composite network impedance model considering the calculation efficiency and model accuracy. For details, please refer to Figure 3 , the network impedance model containing the target device and all associated receiving devices includes the following steps:

[0074] S311: Dividing the preset power grid range into different regions according to the electrical distance with the target device as the center, the different regions including a core region, a buffer region and an external region; this step represents that the preset power grid range is divided into three concentric regions with the target device as the center (the center of the circle), according to the electrical association strength of the node impedance value and the target device: the core region is composed of devices with impedance values lower than a first threshold value and directly closely connected with the target device; the buffer region is composed of devices with impedance values between the first threshold value and a second threshold value (the second threshold value is higher than the first threshold value) and with medium electrical coupling degree; and the external region is formed by devices with impedance values higher than the second threshold value and with weak electrical association degree.

[0075] S312: Constructing the network impedance model for the corresponding devices in different regions, that is, the network impedance model is constructed by differentiated modeling in different regions:

[0076] First, for the core area, construct an accurate model that includes the impedance coefficients of all devices within the area (e.g., establish an accurate π-type equivalent model containing complete impedance parameters for all devices in the core area); second, for the buffer zone, construct a simplified model formed by merging secondary loads (e.g., use load merging simplification techniques to form an aggregated load model for buffer zone devices); third, for the external area, perform external network equivalence steps to construct an equivalent model (e.g., use the WARD equivalence method to establish an equivalent power source-load composite model for the external area network).

[0077] S313: Integrate the precise model, simplified model, and equivalent model at the electrical boundary nodes of the preset power grid range to form the network impedance model; this step means that the precise model, simplified model, and equivalent model are finally integrated into a unified calculation composite network impedance model through node docking.

[0078] The above technical solution divides the preset power grid area into a core area, a buffer area, and an external area based on electrical distance partitioning. Differentiated modeling strategies—accurate, simplified, and equivalent models—are employed for each area. Finally, a composite network impedance model is formed through node docking and integration, achieving the goal of significantly improving the overall network power flow calculation efficiency while maintaining the calculation accuracy of the core area. Furthermore, to ensure the efficiency of power flow projection calculations while fully considering the potential impact of the external power grid on fault handling strategies within the preset area, and to avoid ignoring boundary electrical coupling effects due to simplified modeling, this embodiment employs the following technical means to construct the equivalent model for the external area. Specifically, the steps for constructing the equivalent model for the external area include the following:

[0079] The electrical boundary nodes between the buffer zone and the external zone are determined, and the equivalent injected power at these boundary nodes is calculated. This equivalent injected power is used as the boundary tie-line power result (the boundary tie-line power result and the projected power value together constitute a security assessment dataset for candidate disposal strategies). By calculating the equivalent impedance from the boundary nodes to the external system, an equivalent model that can characterize the dynamic response characteristics of the external network is formed. This technical solution combines traditional network equivalent techniques used for static analysis with the needs of dynamic strategy security assessment. Through the collaborative analysis of the boundary tie-line power result and the projected power values ​​of internal equipment, a complete multi-dimensional security assessment dataset is constructed. This approach preserves the electrical influence characteristics of the external power grid while improving the practicality and reliability of power grid simulation calculations.

[0080] This embodiment also provides an adaptive logic analysis system 500 for power grid heavy overload faults. Please refer to [link to relevant documentation]. Figure 4A modular schematic diagram of the adaptive logic analysis system 500 for power grid overload fault is shown, which is mainly used for dividing the functional modules of the adaptive logic analysis system 500 for power grid overload fault according to the above-mentioned embodiments of the method. For example, each functional module can be divided, or two or more functions can be integrated into one processing module. The above-mentioned integrated module can be realized in the form of hardware or in the form of a software functional module. It should be noted that the division of the module in the present application is schematic, and is only a logical functional division. When each functional module is divided according to each function, for example, Figure 4 Only a system / device schematic diagram is shown, wherein the adaptive logic analysis system 500 for power grid overload fault can include a first analysis unit 510, a second analysis unit 520, a first processing unit 530 and a second processing unit 540. The functions of each unit module are described below.

[0081] The first analysis unit 510 is used to collect a multi-source data set of a target device in real time, identify the overload condition of the target device and the reason type of the overload based on the multi-source data set; wherein the multi-source data set includes electrical quantity data, physical quantity data, environmental quantity data and power grid topology data; the reason type includes natural load growth, fault flow transfer and environmental change; in some embodiments, the first analysis unit 510 is also used to divide a preset power grid range into different regions according to electrical distance with the target device as the center, the different regions including a core region, a buffer region and an external region; a network impedance model is constructed for corresponding devices located in different regions: for the core region, an accurate model containing impedance coefficients of all devices in the region is constructed; for the buffer region, a simplified model formed by merging secondary loads is constructed; for the external region, an external network equivalence step is performed to construct an equivalent model; the accurate model, the simplified model and the equivalent model are integrated at the electrical boundary nodes of the preset power grid range to form the network impedance model. And used to determine the electrical boundary nodes between the buffer region and the external region, calculate the equivalent injection power at the boundary nodes, which is the boundary tie-line power result; by calculating the equivalent impedance from the boundary nodes to the external system, the equivalent model of the external region is formed.

[0082] The second analysis unit 520 is used to calculate the real-time thermal stability margin of the target device based on the multi-source data set, determine the remaining safety time of the target device based on the real-time thermal stability margin, and judge the thermal stability state of the target device according to the remaining safety time;

[0083] The first processing unit 530 is configured to generate at least one candidate treatment strategy in combination with the target device overload reason type and thermal stability state, perform preset power grid range power flow calculation deduction for each candidate treatment strategy, and predict new load rate results of the preset power grid range after executing the candidate treatment strategy. When performing the preset power grid range power flow calculation deduction for each candidate treatment strategy, the load rates of the target device and all associated receiving devices after load transfer are calculated, and the new load rate results of the preset power grid range are synthesized in combination with all the load rates. In some embodiments, the first processing unit 530 is further configured to construct a network impedance model containing the target device and all associated receiving devices, execute the corresponding candidate treatment strategy in the network impedance model, and generate a deduction network topology. Based on the deduction network topology, load data of the current preset power grid range is subjected to power flow calculation deduction to obtain deduction power values of the target device and all associated receiving devices, and the load rates are calculated in combination with the respective deduction power values and dynamic allowable capacities of the target device and all associated receiving devices. The first processing unit 530 is also configured to collect real-time environmental quantity data of the target device and all associated receiving devices, and calculate the respective dynamic allowable capacities of the target device and all associated receiving devices in combination with the thermal stability model and the real-time environmental quantity data.

[0084] The first processing unit 530 is also configured to determine all monitoring points covered in the upstream and downstream intervals of the corresponding device, screen out abnormal monitoring points in all the monitoring points to obtain an effective monitoring point set, select an acquisition mode based on the distribution form of the effective monitoring point set to obtain respective environmental parameters, obtain spatial continuity parameters of all the monitoring points, calculate the effectiveness confidence of each monitoring point according to the spatial continuity parameters of the monitoring point, obtain parameter variation time series of all the monitoring points, compare and analyze the variation trend of each monitoring point, and obtain a variation trend abnormal value of each monitoring point. The first processing unit 530 is also configured to determine whether to screen out in combination with the variation trend abnormal value and the effectiveness confidence of each monitoring point.

[0085] The preset time window is determined, all parameter change time sequences are intercepted with the same length based on the preset time window, and respective parameter comparison time sequences are obtained; similarity measurement is performed on all parameter comparison time sequences, similarity measurement results are obtained, clustering analysis is performed based on the similarity measurement results, and a change trend feature is obtained; each group of parameter change time sequences is compared with the change trend feature, and a change trend abnormal value is obtained. The abnormal change point position in the parameter change time sequence of each monitoring point is obtained, the basic time window value is determined according to the corresponding abnormal change point position; the parameter change rate in the parameter change time sequence of each monitoring point is obtained, and the window floating value is determined according to the change rate; the individual time window value is obtained based on the basic time window value and the window floating value, and the preset time window is determined according to the distribution of all individual time window values. The information of the common monitoring point in the target device and each associated receiving device is determined, the monitoring point of the common monitoring point type is determined, and the value of the collection proportion in the collection mode is improved.

[0086] The second processing unit 540 is configured to filter all candidate processing strategies based on the new load rate result to obtain an optimal strategy set.

[0087] In the above embodiments, the specific working process of each functional unit can refer to the corresponding content disclosed in the foregoing method embodiments. In addition, each functional unit can be implemented by software, hardware, firmware, or any combination thereof, in whole or in part. When implemented by software, the computer program product can be implemented in whole or in part. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center through a wired or wireless manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)), etc.

[0088] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0089] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0090] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0091] Obviously, persons having ordinary skill in the art can be various modifications and variations for the embodiments of the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the embodiments of the present application belong to the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. An adaptive logic analysis method for power grid heavy overload faults, characterized in that, Includes the following steps: S100: Real-time acquisition of multi-source data sets of the target equipment, and identification of the heavy overload status and cause type of the heavy overload of the target equipment based on the multi-source data sets; wherein, the multi-source data sets include electrical quantity data, physical quantity data, environmental quantity data, and power grid topology data; the cause types include natural load increase, fault power flow transfer, and environmental changes; S200: Calculate the real-time thermal stability margin of the target device based on the multi-source data set, determine the remaining safety time of the target device based on the real-time thermal stability margin, and judge the thermal stability status of the target device based on the remaining safety time; S300: Generate at least one candidate handling strategy by combining the cause type and thermal stability state of the target equipment overload, perform power flow calculation and deduction for each candidate handling strategy within a preset power grid range, and predict the new load rate result of the preset power grid range after the candidate handling strategy is executed; S400: Based on the new load rate results, all candidate processing strategies are filtered to obtain the optimal strategy set; Specifically, when performing power flow calculation and simulation for each candidate disposal strategy within a preset power grid range, the load rate of the target device and all associated receiving devices after load transfer is calculated, and a new load rate result for the preset power grid range is synthesized by combining all load rates. When performing power flow calculation and simulation within a preset power grid range for each candidate disposal strategy, the calculation of the load rate of the target device and all associated receiving devices after load transfer includes the following steps: A network impedance model containing the target device and all associated receiving devices is constructed. The corresponding candidate disposal strategy is executed in the network impedance model to generate a deduced network topology. Based on the deduced network topology, the load data of the current preset power grid range is calculated and deduced to obtain the deduced power values ​​of the target device and all associated receiving devices. The load factor is calculated by combining the deduced power values ​​of the target device and all associated receiving devices and the dynamic allowable capacity. The construction of a network impedance model that includes the target device and all associated receiving devices includes the following steps: Centered on the target device, the pre-defined power grid range is divided into different regions based on electrical distance. These regions include a core region, a buffer region, and an outer region. Network impedance models are constructed for the corresponding devices located in different regions. For the core area, construct an accurate model that includes the impedance coefficients of all devices within that area; For the buffer, a simplified model is constructed by merging minor loads; For the external region, perform external network equivalence steps to construct an equivalence model; The precise model, simplified model, and equivalent model are integrated at the electrical boundary nodes of the preset power grid range to form the network impedance model.

2. The adaptive logic analysis method for power grid overload faults according to claim 1, characterized in that, Obtaining the dynamic allowable capacity of the target device and all associated receiving devices includes the following steps: Real-time environmental data of the target device and all associated receiving devices are collected, and the dynamic allowable capacity of the target device and all associated receiving devices is calculated by combining the thermal stability model and the real-time environmental data. The real-time environmental data refers to the environmental parameters obtained from multiple monitoring points along the corresponding device layout, and the results are obtained by comprehensive calculation based on the environmental parameters of each monitoring point. The environmental parameters include temperature, wind speed and solar radiation intensity.

3. The adaptive logic analysis method for power grid heavy overload faults according to claim 2, characterized in that, The environmental parameters obtained from multiple monitoring points along the corresponding equipment layout include the following steps: Identify all monitoring points in the upstream and downstream sections of the corresponding equipment, and filter out abnormal monitoring points to obtain a set of effective monitoring points. Select a collection mode based on the distribution of the effective monitoring point set to obtain the environmental parameters for each mode. The collection mode includes the collection object, collection range, and collection ratio. The process of screening out abnormal monitoring points from all monitoring points includes the following steps: The spatial continuity parameters of all monitoring points are obtained, and the validity confidence level of each monitoring point is calculated based on the spatial continuity parameters of each monitoring point. The time series of parameter changes of all monitoring points are obtained, and the time series of parameter changes of each monitoring point are compared and analyzed to obtain the outlier values ​​of the change trend of each monitoring point. The outlier values ​​of the change trend of each monitoring point and the validity confidence level are combined to determine whether to screen out the outliers.

4. The adaptive logic analysis method for power grid heavy overload faults according to claim 3, characterized in that, The step of comparing and analyzing the time series of parameter changes at each monitoring point to obtain the outlier values ​​of the change trend at each monitoring point includes the following steps: A preset time window is determined, and all parameter change time series are truncated to the same length based on the preset time window to obtain their respective parameter comparison time series; similarity measurement is performed on all parameter comparison time series to obtain similarity measurement results, and cluster analysis is performed based on the similarity measurement results to obtain change trend characteristics; the difference between each group of parameter change time series and change trend characteristics is compared to obtain change trend outliers.

5. The adaptive logic analysis method for power grid heavy overload faults according to claim 4, characterized in that, Determining the preset time window includes the following steps: Obtain the location of abnormal change points in the parameter change time series of each monitoring point, and determine the basic time window value based on the location of the corresponding abnormal change point; obtain the parameter change rate in the parameter change time series of each monitoring point, and determine the window fluctuation value based on the change rate; obtain the individual time window value based on the basic time window value and the window fluctuation value, and determine the preset time window based on the distribution of all individual time window values.

6. The adaptive logic analysis method for power grid heavy overload faults according to claim 3, characterized in that, After selecting the acquisition mode based on the distribution of the effective monitoring point set and obtaining the respective environmental parameters, the process further includes the following steps: Determine the information of common monitoring points in the target device and each associated receiving device, identify monitoring points that are classified as common monitoring points, and increase their acquisition ratio in the acquisition mode; wherein, the common monitoring point refers to the monitoring point used for data acquisition by at least two devices in the target device and all associated receiving devices.

7. The adaptive logic analysis method for power grid heavy overload faults according to claim 1, characterized in that, The step of performing external network equivalence steps to construct an equivalence model for the external region includes the following steps: The electrical boundary nodes between the buffer zone and the external zone are determined, and the equivalent injected power on the boundary nodes is calculated. This equivalent injected power is used as the boundary tie line power result. The equivalent impedance of the external system as seen from the boundary nodes is calculated to form an equivalent model of the external zone. The boundary tie line power result and the inferred power value together constitute a security assessment dataset for screening candidate disposal strategies.

8. An adaptive logic analysis system for power grid heavy overload faults, characterized in that, include: The first analysis unit is used to collect multi-source data sets of the target equipment in real time, and identify the heavy overload situation and cause type of the heavy overload of the target equipment based on the multi-source data sets; wherein, the multi-source data sets include electrical quantity data, physical quantity data, environmental quantity data and power grid topology data; the cause types include natural load increase, fault power flow transfer and environmental changes; The second analysis unit is used to calculate the real-time thermal stability margin of the target device based on the multi-source data set, determine the remaining safety time of the target device based on the real-time thermal stability margin, and judge the thermal stability status of the target device based on the remaining safety time. The first processing unit is used to generate at least one candidate disposal strategy by combining the cause type and thermal stability state of the target device overload, and to perform power flow calculation and simulation for each candidate disposal strategy within a preset power grid range, and to predict the new load rate result of the preset power grid range after the candidate disposal strategy is executed; wherein, when performing power flow calculation and simulation for each candidate disposal strategy within a preset power grid range, the load rate of the target device and all associated receiving devices after load transfer is calculated, and the new load rate result of the preset power grid range is synthesized by combining all load rates; When performing power flow calculation and simulation within a preset power grid range for each candidate disposal strategy, the calculation of the load rate of the target device and all associated receiving devices after load transfer includes the following steps: A network impedance model containing the target device and all associated receiving devices is constructed. The corresponding candidate disposal strategy is executed in the network impedance model to generate a deduced network topology. Based on the deduced network topology, the load data of the current preset power grid range is calculated and deduced to obtain the deduced power values ​​of the target device and all associated receiving devices. The load factor is calculated by combining the deduced power values ​​of the target device and all associated receiving devices and the dynamic allowable capacity. The construction of a network impedance model that includes the target device and all associated receiving devices includes the following steps: Centered on the target device, the pre-defined power grid range is divided into different regions based on electrical distance. These regions include a core region, a buffer region, and an outer region. Network impedance models are constructed for the corresponding devices located in different regions. For the core area, construct an accurate model that includes the impedance coefficients of all devices within that area; For the buffer, a simplified model is constructed by merging minor loads; For the external region, perform external network equivalence steps to construct an equivalence model; The precise model, simplified model, and equivalent model are integrated at the electrical boundary nodes of the preset power grid range to form the network impedance model; The second processing unit is used to filter all candidate processing strategies based on the new load rate result to obtain the optimal strategy set.

Images