CN105245015B - Multi-Agent-Based Hierarchically Extended Power Grid Fault Information Processing System and Method - Google Patents

Abstract

The invention discloses a kind of delamination electric network fault information processing system based on many AGENT, including: Agent system, described Agent system communicates respectively with guarantor's communication system of the electrical network out-of-limit edge device of poor stream, event recording system and SCADA system；Described Agent system is by gathering the differential partition boundaries point electric current of each electrical network, after carrying out data normalization process, carry out the differential current computing with equipment for minimum division unit, when having monitored unit difference stream more in limited time, start dependent failure profiler, carry out the judgement of fault zone and position of failure point.The method have the benefit that difference this dimension of stream by introducing region fault current, by coming from the different aforementioned sources information fusion of different substation end, upwards can extend to the judgement in trans-regional cascading failure region, can refine downwards and the accurate range finding of trouble point on Judgement of failure, circuit.

Description

Multi-Agent-Based Hierarchically Extended Power Grid Fault Information Processing System and Method

technical field

The invention relates to a power grid fault information fusion processing system, in particular to a multi-AGENT-based layered expansion power grid fault information processing system and method.

technical background

Protection engineers often need to extract, analyze and synthesize data from a series of monitoring equipment, with the purpose of diagnosing power system faults or disturbances, and determining whether the relevant protection actions are correct. Through this analysis, confirm the protection malfunction and faulty equipment, and at the same time complete the action analysis of the relevant automatic devices and the action value check of the protection setting value.

A series of effective protection analysis tools have been developed using data from fault recorders (DFR) for fault or disturbance diagnosis tasks, such as alarm interpretation systems, fault classification and protection behavior evaluation systems. Although these independent intelligent systems can assist and diagnose in specific aspects, the completion of the entire work still requires human intervention to compare and interpret the generated information.

Once a fault occurs, generating a large amount of DFR data will cause many problems. The first is the problem of data overload; many line fault recording devices are not the point of failure, and the interference caused by the fault may cause the voltage to drop and start the wave recording and upload a large amount of irrelevant information; in addition, it will lead to the problem of data loss; when the fault record is passed The cache is automatically rolled for recording. When the engineer extracts the fault data, a large amount of data may have been cached in the fault recording file. If a transmission line experiences a series of large disturbances in a short period of time, the fault file may be overwritten by buffer rollover before extraction for analysis.

The existing guarantee system can analyze the whole process of protection actions from completely different angles. Therefore, if an automatic fault diagnosis comprehensive system can automatically complete timely integration including data collection and direct related information, and present it to protection engineers, and realize preliminary automatic analysis at the same time, this system will provide protection and on-site dispatching personnel with great benefits. s help.

According to the experience of control personnel, a large amount of data will always flow in before and after the fault, forming a big digital storm. For example, for a 24-hour fault handling process, the power monitoring system can generate more than 15,000 alarms and hundreds of fault records.

Generally, protection engineers start to evaluate the protection action behavior from the analysis of SCADA data, which can enable the engineer to initially identify the occurrence of power system interference and the line of interference; verify some aspects of the protection action behavior, such as: lack of protection or trip alarm , some protection actions are unsuccessful.

But the disadvantage of this method is that the protection analysis is not intuitive enough, and sometimes it is difficult to analyze because of the sequence of protection actions, especially for the judgment of complex faults.

Therefore, how to start an automatic fault analysis program by a specific protection action characteristic in an intuitive way, and close this automatic analysis program with the isolation of the fault, and complete a cycle of analysis is very important. The advantage of this is that irrelevant information can be automatically shielded, and an event can be analyzed independently, which is conducive to the step-by-step interpretation of sequence events.

Contents of the invention

The purpose of the present invention is to solve the above problems, and propose a multi-agent-based hierarchical extended power grid fault information processing system and method. The system uses wide-area fault recording information to perform differential current calculation of fault currents in related areas to realize fault areas. Accurately identify, and display the fault current and diagnosis results of each interval on the station map uniformly and visually. With the organic coupling guarantee system, record system and SCADA system, etc., the monitoring of fault operation status, rapid identification of fault areas and precise positioning of fault points are realized.

In order to achieve the above object, the present invention adopts the following technical solutions:

A multi-Agent-based layered and extended power grid fault information processing system, comprising: a multi-Agent system, the multi-Agent system communicates with a credit guarantee system, a story system and a SCADA system of a power grid differential current cross-limit boundary device; The Agent system collects the current at the boundary points of the differential partitions of each power grid, and after normalizing the data, calculates the differential current with the equipment as the smallest division unit. When it detects that the differential current of an equipment unit exceeds the limit, it starts the relevant fault analysis and diagnosis program , to determine the fault area and the location of the fault point.

Described multi-agent system comprises:

Intelligent data collector: used to collect current signals at the boundary points of each differential partition, and perform differential current calculation with equipment as a unit;

Data synchronization coaxial processing agent: extract the relevant fault recorder ID and report path, standardize the reports from different data sources, and realize the synchronization and normalization of the sampled data;

Interpolation analysis agent: use the interpolation method to unify different sampling frequencies into the same sampling frequency, compress and display the waveform data without sampling points in the form of an envelope line, and only record the compressed storage part of the effective value in the wave recording data;

Fault location agent: used to determine the location of line fault points;

Protection action sequence analysis agent: use the reasoning protection model library based on the differential current model to identify whether the protection components are operating correctly by comparing the behavior of the actual relay with the behavior obtained from the reasoning protection model library.

The data synchronous coaxial processing agent, the difference analysis agent, the fault distance measurement agent and the protection action sequence analysis agent in the multi-agent system communicate with each other.

Each agent in the multi-agent system is an expert processing system with corresponding diagnosis knowledge distributed.

The fault location agent uses the zero-sequence quantity as the state quantity to determine and distance the fault point.

A multi-agent-based method for hierarchically expanding power grid fault information processing system. When the differential current calculation is abnormal, the minimum protection identification area is determined by the wide-area differential principle, and the effective protection zone with the failed element as the boundary point is determined by the near-far backup principle. Expand the maximum identification area of protection to judge the fault area and the location of the fault point;

The specific method is as follows:

(1) Real-time collection of data from the credit guarantee system, record system and SCADA system of the boundary equipment of the differential current of the power grid, normalize the data from different data sources, and set the threshold value of the differential current;

(2) Calculate the differential flow with the equipment as the smallest division unit, and judge whether the differential flow crosses the line, if so, start the multi-agent system to retrieve the relevant fault records, and go to the next step, otherwise, return to step (1);

(3) Determine whether the differential current exceeds the limit caused by data collection, if so, determine the correction plan, and perform background or loop correction; otherwise, go to the next step;

(4) Calculate whether the differential current of the adjacent components crosses the line, if yes, it is judged to be a fault in the intersection area of the transmission line, and enter the next step; otherwise, directly transfer to the next step;

(5) After the differential protection device detects a fault, it sends fault protection information according to the reasoning protection model library based on the differential current model, and judges whether the fault protection information is consistent with the action of the isolating switch. If yes, return to step (1); otherwise, judge In order to protect maloperation, restore power transmission in time, perform closing, and update the information related to protection maloperation to the reasoning protection model library based on differential current model.

After determining the fault in the crossing area of the transmission line, according to the operation mode of the power grid before the line fault, use the data recorded in the fault recorder to search for the fault point. If a fault point is found in the fault line area, and the lines on both sides are zero If the ratio of the sequence current is equal to the ratio of the line impedance on both sides, then the fault point is determined to be the desired fault point.

The differential current exceeding the limit caused by data acquisition includes: the polarity of the fault recorder, the phase error, or the secondary circuit multi-point grounding.

Beneficial effects of the present invention:

By introducing the dimension of differential current of regional fault current, it can be solved by locking the judgment of faulty equipment in the two-dimensional space of time and faulty equipment. At the same time, using this as the event granularity standard, through the fusion of different information sources from different substations, it can be extended upwards to the judgment of cross-regional cascading fault areas, and downwards can be refined and judged by the nature of faults and accurate measurement of fault points on the line. distance. Since both information fusion and the bottom line fault location are based on the integration, classification, and calculation from different real-time data sources, that is to say, information fusion and graphic normalization processing technology can be used to show the fault behind the fault to the greatest extent. Reasons and deep-seated problems, to a large extent, realize the intelligent decision-making auxiliary tools required for automatic analysis-type scheduling, and provide a favorable tool for dispatchers to quickly analyze.

Description of drawings

Figure 1 is a fusion direction diagram of information corresponding to MAS system information and primary grid diagram;

Figure 2 is a diagram of the open structure of MAS;

Fig. 3 is a flowchart of fault handling in the present invention;

Fig. 4 is a schematic diagram of merging fault data from different information sources;

Figure 5 is a zero-sequence equivalent network diagram;

Figure 6 is a flowchart of distance measurement.

detailed description

Below in conjunction with accompanying drawing and specific embodiment the technical scheme of the present invention is further described:

As shown in Figure 1, each component of the power grid has a differential protection device. When the differential current value in any differential current calculation area exceeds the limit, the diagnostic system will be activated to centrally retrieve relevant fault records. Including the security system, record system and SCADA system related to the boundary equipment of the differential flow. The retrieved detailed information is used to evaluate protection action behavior and fault diagnosis. At the same time, as an analysis tool, the system integrates a reasoning protection model library based on the differential current model (DIFFERENTIAL-CURRENTMODELBASEDREASONGING, DMBR). Correct operation of the protective components can be identified afterwards by comparing the actual relay behavior with that predicted by the model.

However, data stored within DFR must be converted from the COMTRADE format. At the same time, since the information about the differential flow crossing the boundary may belong to different substations in physical location, data synchronization is required and presented on the same coordinate time axis. With the above preparations, relevant information can be integrated and fault diagnosis can be carried out in real time. The specific process is shown in Figure 1.

1 The framework of the comprehensive solution platform for fault diagnosis based on MAS

It can be seen from the above analysis that remote fault diagnosis can meet the reliability and security requirements and development trends of power system operation. This paper defines and proposes a web-based multi-agent system collaborative diagnosis platform for power systems. In the architecture design framework of MAS, each agent has its specific role assignment, and they cooperate with each other to complete the diagnosis process. The platform itself can support the communication and information interaction among its constituent agents.

1.1MAS system architecture

The application of common collaborative diagnosis system is an open system. From an architectural point of view, the system provides a framework for coordinating several specific agent behaviors. As shown in Figure 1, the agents in the system include a series of diagnostic agents (including difference analysis agent, fault distance measurement agent, and protection action sequence analysis agent) working on the server side and data synchronization coaxial processing agents. The diagnostic knowledge is distributed among multiple agents, and each agent is an expert processing system. Whether the whole system can effectively solve the problem depends on the accumulated professional knowledge of the agent community. In addition to using the intelligent data collector to monitor the agent on the client side, other agents collect current through the boundary points of each differential partition, and perform differential current calculation with the device as a unit. The method for calculating the differential flow may adopt an existing method. When it is detected that the differential flow of an equipment unit exceeds the limit, the relevant fault analysis and diagnosis program is started. The monitoring agent of the client sends a request to the server of the diagnosis/treatment service, and the server sends the response and the diagnosis result back to the client. The specific structure is very simple. The flow of information is shown in Figure 2.

It can be seen from the illustrated platform architecture that this platform is an open platform, which is reflected in two aspects: from the perspective of platform composition, since the platform is coupled with multiple technologies, and the growth of each technology itself is It constitutes the growth of the overall level of the platform, so it can be considered as the integration of the vertical software level; from the perspective of physical area, as shown in Figure 1, from the division of the area with the line object as the smallest unit to the judgment of the protection area based on the initial judgment of the fault In the fault analysis of the area extension of the minimum unit starting point, the physical area is expanded layer by layer. Of course, this kind of expansion will not cause additional costs, because this platform is built on the level of the storybook and credit guarantee system that has already been built, and what the system does is only information fusion and integrated analysis. On the other hand, it will not cause a sharp increase in the amount of information processing and an increase in the "dimension" of analysis. Because the granularity of the analysis is the differential flow area with the equipment as the minimum judgment unit, and each "step" of expansion starts from the failure point of the "action" and expands with the equipment as the unit. Therefore, from the perspective of analysis, Speaking is also a "dimension" and does not increase the "dimension" of analysis.

1.2 Information-limited extension method based on the starting point of the action "failure" element

A.A4 protection action failed

As shown in Figure 1, when the A4 differential protection action fails, the calculation of the differential flow is abnormal. At this time, the judgment conclusion based on the line L2 as the minimum protection area fails, so the information collection and judgment area is expanded starting from the failure action boundary point A4, Put L2 and B7 in the list of possible fault objects.

Start the secondary information expansion collection. From the information collected by the Baoxin substation system, it can be seen that the intersecting area of the distance protection actions of A3 and A4 is L2. After a period of information integration, the faulty device can be locked and the evaluation of the protection action behavior is given. It can be seen that the differential current calculation in the "information closure" mode is used in the normal fault component locking judgment, and when the judgment fails, the "semi-open" distance protection is limitedly integrated, and the distance protection action area information is closed , to lock the faulty device again.

B. Switch 4 failure

The analysis is still based on the same case. When the differential current calculation based on L2 exceeds the limit and is issued by the differential protection action command, the result is that the switch A3 trips, and A4 fails to trip. At this time, the starting bus differential fails. The protection jumper is connected to the A5 switch connected to B7. Apparently, this analysis process comes from the information synthesis and synchronization of differential flow over-limit analysis and credit guarantee system. It can be seen that this comprehensive analysis system not only comes from the synchronization of physical sampling, but also includes the cross-platform generalized information synchronization of SCADA system, credit system and old record physical quantity collection.

From the above analysis, it can be seen that whether it is the primary locking of the faulty device based on whether the differential current exceeds the limit, or the secondary judgment based on the limited expansion of the information of the action failure object, its essence is the information synchronization and fusion of the story, guarantee and SCADA system. But this kind of fusion has a clear direction and directivity. The expansion of its physical area inherits the idea of protecting near and far backups. Taking A3 as an example, its minimum protection area is the line L2 and bus B8 as its boundary point; with its far backup protection range as the boundary, the largest protection area Defined as lines L1, L3, L5 and bus B7. For example, when protection A3 has no information output, its information blind spot can still locate the fault in L2 by using the extended maximum protection zone boundary points A1 and A10 as the intersection of L2 far backup protection range and A4 protection zone. The specific process is shown in Figure 3 as follows.

It can be seen from the fault processing flow in Figure 3 that the fault diagnosis process is essentially a simulation of the accident analysis thinking of the protection project, because there are definitions of the minimum and maximum protection areas, and at the same time, there is a differential flow that is simultaneously expanded with the protection area. Intuitive fault criterion makes the division of events clearly defined. Therefore, there are effective tools for analyzing complex faults, sequence trips and system abnormal conditions. Because the threshold value of differential flow is the most effective criterion for distinguishing faults from abnormal conditions. However, the differential flow partition calculation method with equipment as the smallest division unit can effectively lock the non-unique equipment that fails at the same time. The appearance and disappearance of the differential flow can clearly reflect an independent event package of a sequence protection action.

2 Main key technologies of the system

2.1 Sampling Synchronization and Frequency Normalization AGENT Based on Interpolation Method

Figure 4 shows the schematic diagram of the system for standardization processing and merging of graphics and related information from different information sources (including waveforms from protection and old records). It can be seen that the fault protection format from the protection device is COMTRADE, which is different from the format type of the fault recorder, and the report formats from different fault recorders are also different, but no matter which format, the relevant fault recorder can always be proposed. oscilloscope ID and report path, so reports from different data sources can be standardized as long as relevant feature quantities are proposed; at the same time, in order to facilitate in-depth analysis, different protection principles can also be set through different differential action zone settings, so that Realize the analysis of the protection action behavior when the actual fault occurs. The middle part details the normalization and merging process of the waveform diagram. The right side shows the protection action behavior analysis and waveform and information fusion results when the actual fault occurs after the corresponding action area is configured, which can facilitate the judgment of the fault nature and protection behavior. Calibration, the above process is completed automatically through pre-setting.

2.2 High-impedance ground fault location AGNET based on zero-sequence network based on story information

What needs to be pointed out is that this system adopts the interpolation method to unify different sampling frequencies into the same sampling frequency, compresses and displays the waveform data without sampling points in the form of an envelope line, and at the same time only records the compressed storage part of the effective value in the wave recording data, and Treat accordingly. The normalized wave form in this way can be used to calculate the vector size in real time through the Fourier algorithm, and then calculate the differential flow value with the device as a unit, and display it in real time in the primary diagram of the device in the form of intuitive graphics.

At present, the distance measurement method of the fault system generally adopts the algorithm of solving differential equations, but because it is based on the sampling value of a single local device, the number of equations is limited and only two unknowns can be solved, so it is assumed that the current at the protection installation is in the same phase as the fault current , while omitting to solve the state quantity of fault current angle. However, this assumption cannot be satisfied when a high-resistance grounding fault is caused, because the fault current angle cannot be considered to be the same angle as the current at the protection installation due to the existence of transition resistance. And because the wide-area multi-terminal sampling is adopted, the grounding impedance value allowed by the project and the corresponding fault point location can be screened out by the back substitution method, as shown in Figure 6.

Considering it is a high-resistance grounding fault, so the zero-sequence quantity is selected as the state quantity. The zero-sequence equivalent network is shown in Figure 5, and its state equation is

$\frac{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}\mathrm{<span\; class="notranslate">\; N</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}}{{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

It can be seen that the state equation has nothing to do with the resistance value of the grounding resistance Rg. The zero-sequence current of the two stations can be obtained from the protection recording files of the protection devices on both sides.

After calculating the wide-area differential current based on the real-time information of the history record, after determining that the fault area is the transmission line, and then according to the operation mode of the power grid before the line fault, use the data recorded in the fault recorder to search with the calculation program module in it, and find the faulty line. For a fault point, the ratio of the zero-sequence current on both sides of the line is equal to formula (1), so this fault point is the desired fault point.

3 case analysis

Case: Thunderstorm weather in a certain area, a certain 220 kV line tripped, through the above-mentioned system and method, the integrated interface based on the MAS fault diagnosis system integration history, credit, and SOE information of the SCAD system can be given; in addition, in the fault In the report, the discrimination of the fault point and the judgment of the nature of the fault are given, corresponding to which is the wave record of the fault quantity. In the analysis of the action situation, the analysis of the action behavior of the differential main protection is mainly given. The corresponding fault recording window is just a button-like interface, which gives the synthetic differential current value. If further analysis is carried out, the corresponding interface can be opened as needed to obtain detailed wave recording information.

4 Conclusion

This paper introduces a complete set of solutions for rapid handling of power grid faults and accidents based on MAS. It decomposes the complex problem of power grid faults layer by layer with appropriate dimensions and granularity, uses normalization technology to integrate existing systems, and uses MAS open, multi- Hierarchical and growth-oriented framework, effectively integrating the existing storybooks, guarantees, and SCADA systems, and determining the minimum protection identification area through the principle of wide-area differential, and determining the maximum identification of effective extended protection with failed components as boundary points based on the principle of near and far backup The advantage of this approach is that by introducing the dimension of the differential current of the regional fault current, it can be solved by locking the judgment of the faulty equipment in the two-dimensional space of time and faulty equipment. Tongyuanshi takes this as the event granularity standard, and through the fusion of different information sources from different substations, it can be extended upward to the judgment of cross-regional cascading fault areas, and downward can be refined and judged by the nature of faults and fault points on the line. Accurate ranging. Since both information fusion and the bottom line fault location are based on the integration, classification, and calculation from different real-time data sources, that is to say, information fusion and graphic normalization processing technology can be used to show the fault behind the fault to the greatest extent. Reasons and deep-seated problems, to a large extent, realize the intelligent decision-making auxiliary tools required for automatic analysis-type scheduling, and provide a favorable tool for dispatchers to quickly analyze.

Although the specific implementation of the present invention has been described above in conjunction with the accompanying drawings, it does not limit the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, those skilled in the art do not need to pay creative work Various modifications or variations that can be made are still within the protection scope of the present invention.

Claims (7)

1., based on a delamination electric network fault information processing system of many AGENT, it is characterized in that, including: Agent system, described Agent system communicates respectively with guarantor's communication system of the electrical network out-of-limit edge device of poor stream, event recording system and SCADA system；Described Agent system is by gathering the differential partition boundaries point electric current of each electrical network, after carrying out data normalization process, carry out the differential current computing with equipment for minimum division unit, when having monitored unit difference stream more in limited time, carry out the judgement of fault zone and position of failure point；

Described Agent system includes:

Intelligent data acquisition unit: for gathering the current signal of each differential partition boundaries point, carry out the differential current computing being unit with equipment；

Data syn-chronization coaxially processes agent: extract dependent failure oscillograph ID and report path, is standardized the report from different pieces of information source processing, it is achieved therefore record sampling data synchronization and normalization；

Interpolation analysis agent: adopting interpolation method is same sample frequency by the unification of difference sample frequency, the Wave data not having sampled point carries out envelope mode and compresses display, and the compression simultaneously only recording virtual value in recorder data stores part；

Fault localization agent: for the position of measurement line trouble point；

Protection act sequence analysis agent: utilize the reasoning protection model library based on difference flow model, by the behavior of relatively actual relay whether consistent with the behavior drawn by reasoning protection model library come identification protecting assembly whether correct operation.

2. a kind of delamination electric network fault information processing system based on many AGENT as claimed in claim 1; it is characterized in that, the data syn-chronization in described Agent system coaxially processes intercommunication between agent, differential analysis agent, fault localization agent and protection act sequence analysis agent.

3. a kind of delamination electric network fault information processing system based on many AGENT as claimed in claim 1, is characterized in that, in Agent system, each agent is a specialist processing system, and distribution has corresponding diagnostic knowledge.

4. a kind of delamination electric network fault information processing system based on many AGENT as claimed in claim 1, is characterized in that, described fault localization agent carries out determination and the range finding of trouble point using zero sequence amount as quantity of state.

5. the control method of the delamination electric network fault information processing system based on many AGENT as claimed in claim 1; it is characterized in that; when differential current computing occurs abnormal; Minimal Protective cog region is determined by wide area differential principle; determine, with standby principle near, remote, the maximum cog region of effective extended protection being boundary point with failure element, carry out the judgement of fault zone and position of failure point；

Concrete grammar is as follows:

(1) data from different pieces of information source are normalized, and set the threshold value of difference stream by the data protecting communication system, event recording system and SCADA system of the out-of-limit edge device of Real-time Collection electrical network difference stream；

(2) carry out differential current computing with equipment for minimum division unit, it is judged that whether difference stream is out-of-limit, if it is, start the failure logging that Agent system retrieval is relevant, proceed to next step, otherwise, return step (1)；

(3) the difference stream caused when determining whether data acquisition is out-of-limit, if it is, determine correcting scheme, carries out backstage or loop is revised；Otherwise, next step is proceeded to；

(4) whether out-of-limit calculate adjacent elements difference stream, if it is, be judged as transmission line of electricity zone of intersection fault, enter next step；Otherwise, next step is directly proceeded to；

(5), after differential protection detects fault, error protection information is sent according to the reasoning protection model library based on difference flow model, it is judged that whether error protection information is consistent with isolation switch motion, if it is, return step (1)；Otherwise, it is judged that for false protection, recover power transmission in time, perform combined floodgate, and by the information updating relevant to false protection to the reasoning protection model library based on difference flow model.

6. control method as claimed in claim 5, it is characterized in that, after being defined as transmission line of electricity zone of intersection fault, according to power system operating mode before line fault, utilize recorded data in fault oscillograph, carry out trouble point search, if at faulty line range searching a to trouble point, and the ratio of its both sides circuit zero-sequence current is equal with the ratio of both sides line impedance, it is determined that this trouble point is required trouble point.

7. control method as claimed in claim 5, is characterized in that, the difference caused during data acquisition flows out-of-limit situation and includes: the polarity of failure wave-recording, phase error or secondary circuit multipoint earthing.