A switch cabinet comprehensive online monitoring analysis and fault prediction method and system

Abstract

The present invention discloses a comprehensive online monitoring and analysis and fault prediction method and system for switch cabinets, belonging to the technical field of power equipment monitoring. It includes: configuring various types of data acquisition devices according to the functional requirements of the switch cabinet; creating dynamic self-checking standards based on data types to determine whether there is a fault; if there is a fault: issuing a corresponding tripping or alarm instruction, and recording the fault occurrence point to update to a fault information set in order of time points; each time the fault information set is updated, the updated fault information set is processed to obtain comprehensive fault information data, and the comprehensive fault information data is input into a fault prediction model, and the fault prediction model outputs the future time point and the time point when the corresponding fault point fails again, and generates corresponding control decisions based on the future time point. In the invention, the current, temperature, arc light and other aspects of the switch cabinet are monitored; through the algorithm, defect identification and early warning are achieved, and the fault missed detection rate is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of power equipment monitoring, and in particular relates to a comprehensive online monitoring and analysis and fault prediction method and system for a switch cabinet. Background Art

[0002] A PT cabinet exploded in a petrochemical company, causing a section of the busbar to lose power. The main reason for the accident was that the PT had manufacturing quality defects and lacked corresponding trend change monitoring, which led to the hidden danger expanding to equipment failure. A petrochemical company also had a three-phase short circuit fault in a 35kV incoming cabinet, causing a low-voltage power grid power shake accident. The main reason for the accident was the lack of management and monitoring of the operating status of the incoming cabinet, which led to the failure to timely discover the insulation hidden dangers of the switch cabinet, which developed into a three-phase short circuit fault.

[0003] The above cases reflect that the traditional electrical equipment operation and maintenance - manual regular inspection method can no longer adapt to the development of enterprises, mainly highlighting the following problems:

[0004] (1) For sudden and repetitive electrical equipment failures, manual inspections cannot promptly identify the problem and effectively eliminate the fault, resulting in missed inspections.

[0005] (2) Manual inspections do not collect data on the equipment's operating status continuously, and it is impossible to establish an effective electrical status data trend. At the same time, it is impossible to predict the time point of an upcoming failure, which means that decisions can only be made after a failure occurs, which wastes a lot of time. Summary of the invention

[0006] Purpose of the invention: In order to solve the above problems, the present invention provides a method for comprehensive online monitoring and analysis and fault prediction of a switch cabinet, comprising the following steps:

[0007] Step 1: Configure various types of data acquisition devices D according to the functional requirements of the switch cabinet type(i) , type(i) represents the number of the data acquisition device in the data type type;

[0008] Step 2: Create a dynamic self-check standard S based on the data type type type , the data acquisition device D type(i) Real-time data collected type(i) Input dynamic self-test standard S type , determine whether there is a fault; if there is a fault, execute step 3;

[0009] Step 3: Issue the corresponding tripping or alarm command, and record the fault occurrence point and update it to the fault information set F in order of time points. type(i) ; Execute step 4;

[0010] Step 4: Fault information set F type(i) Each time the update is performed, the updated fault information set F type(i) Perform a data processing to obtain comprehensive fault information data F type-t , the fault information comprehensive data F typ Input a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again, and generates a corresponding control decision based on the future time point;

[0011] Step 5: If the actual time point at which the corresponding fault point fails again is earlier than the future time point, update the fault information set F based on the actual fault. type(i) , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set F at the future time point type(i) , repeat step 4.

[0012] In a further embodiment, the data type type includes: current data, illumination data, temperature data, and partial discharge data.

[0013] In a further embodiment, the fault information set F type(i) Further expressed as: F type(i) ={f1,f2,…,f n}, n is the number of failures that have occurred in the data acquisition device within the data type type from the start to the end of the current time point, f n is the fault information of the nth fault, which includes: the fault occurrence point and the fault type, expressed in the form of f n =w q→t , q represents the type of fault, and w represents the weight of fault q at time point t.

[0014] In a further embodiment, the fault information comprehensive data F type- The method of obtaining is as follows: type(i) The elements in are fused in a parallel superposition manner to obtain the fused spectrum {F t}, where F t =f1∪f2∪…∪f n ; Normalize the fused spectrum to get the sum of the spectrum dimensions, F type-t = dim({F t}).

[0015] In a further embodiment, the fault prediction model is expressed as follows:

[0016] Calculate predicted values Among them, fk ∈F type(i) ;

[0017] The next time point L(·) is a differentiable convex function, and Ω(·) is a regularization penalty term.

[0018] In a further embodiment, the dynamic self-checking standard S type The creation process is as follows:

[0019] The corresponding self-check standard is selected according to the current data type. type It includes: one or more of the self-checking standards for current data, self-checking standards for illumination data, self-checking standards for temperature data, and self-checking standards for partial discharge data.

[0020] In a further embodiment, when the current data is current data, the corresponding judgment process is as follows:

[0021] The current data at least includes the following data: I l ,Will I l Corresponding to the fixed value O θ , O f , O l A comparative analysis was conducted and the results are as follows:

[0022]

[0023] in, is the current three-phase current monitored in overcurrent section I, overcurrent section II, and overcurrent section III; θ corresponds to overcurrent section I, overcurrent section II, or overcurrent section III, and i corresponds to phase a, phase b, or phase c;

[0024]

[0025] in, is the current three-phase load current;

[0026]

[0027] Among them, I l is the current zero-sequence current value.

[0028] In a further embodiment, when the current data is illumination data, the corresponding judgment process is as follows: the illumination data includes the current illumination U, and U is compared with the illumination constant U 0 A comparative analysis was conducted and the results are as follows:

[0029]

[0030] In a further embodiment, when the previous data is temperature data, the corresponding judgment process is as follows: the temperature data includes the current temperature value T, T and the temperature constant T 0 A comparative analysis was conducted and the results are as follows:

[0031]

[0032] A switch cabinet comprehensive online monitoring analysis and fault prediction system, used to implement the comprehensive online monitoring analysis and fault prediction method as described above, comprising:

[0033] The first module is configured to configure various types of data acquisition devices D according to the functional requirements of the switch cabinet. type(i) , type(i) represents the number of the data acquisition device in the data type type;

[0034] The second module is configured to create a dynamic self-checking standard S based on the data type type type , the data acquisition device D type(i) Real-time data collected type(i) Input dynamic self-test standard S type , determine whether there is a fault; if there is a fault, execute step 3;

[0035] The third module is configured to issue a corresponding tripping or alarm command and record the fault occurrence point updated to the fault information set F in order of time points. type(i) ; Execute step 4;

[0036] The fourth module is set as the fault information set F type(i) Each time the update is performed, the updated fault information set F type(i) Perform a data processing to obtain comprehensive fault information data F type-t , the fault information comprehensive data F type-t Input a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again, and generates a corresponding control decision based on the future time point;

[0037] The fifth module is configured to update the fault information set F based on the actual fault if the actual time point at which the corresponding fault point fails again is earlier than the future time point. type(i) , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set F at the future time point type(i) , repeat step 4.

[0038] Beneficial effects: By setting up multiple monitoring modules, such as three-stage overcurrent monitoring module, overload monitoring module, zero-sequence overcurrent monitoring module, arc monitoring module, wireless temperature measurement monitoring module, partial discharge monitoring module, smoke sensor module, infrared temperature measurement module; obtaining the current data, light data, temperature data, and partial discharge data of the monitored object, and monitoring the monitored object. Once the acquired data does not meet the predetermined value, it is immediately determined that there is a fault at the corresponding monitoring location, and then tripping or alarming is performed to achieve multi-faceted monitoring. And it has achieved a comprehensive prediction of the time point of the fault, and can make decisions in advance to prepare for emergencies. DETAILED DESCRIPTION

[0039] Example 1

[0040] The present embodiment provides a comprehensive online monitoring, analysis and protection method for switch cabinets (hereinafter referred to as the present method). The implementation of the present method is based on an online power monitoring device, which monitors the medium voltage switch cabinet. The device consists of a current and voltage unit, an arc unit, a CT wireless temperature unit, a smoke sensor unit, a partial discharge unit, an infrared imaging camera unit and a control unit. The above units are communicatively connected with the control unit and can be used in various industrial and mining enterprises and power industries with voltage levels of 110kV and below. The following is a detailed description of the present method:

[0041] A method for comprehensive online monitoring, analysis and fault prediction of a switch cabinet comprises the following steps:

[0042] Step 1: Configure various types of data acquisition devices D according to the functional requirements of the switch cabinet type(i) , type(i) represents the number of the data acquisition device in the data type type; in other words, the control module and the n monitoring modules connected thereto are pre-configured, and any monitoring module includes at least the following working states: put into monitoring state and exit monitoring state

[0043] In order to monitor the medium voltage switchgear in an all-round way, a control module and n monitoring modules are configured, wherein the n monitoring modules at least include: three-stage overcurrent monitoring module, overload monitoring module, zero-sequence overcurrent monitoring module, arc monitoring module, wireless temperature measurement monitoring module, partial discharge monitoring module, smoke sensor module, infrared temperature measurement module. Refers to putting the monitoring module into use. Exit monitoring status It means to exit the monitoring module, indicating that the monitoring module does not participate in monitoring.

[0044] Step 2: Create a dynamic self-check standard S based on the data type type type , the data acquisition device D type(i) Real-time data collected type(i)Input dynamic self-test standard S type , determine whether there is a fault; if there is a fault, execute step three; that is, obtain self-test data based on the control module, and determine whether the self-test data meets the self-test standard: if it meets the self-test standard, it is determined to be a normal operating state, and a first display signal is issued at a predetermined frequency; otherwise, it is determined to be an abnormal operating state, and a second display signal is issued.

[0045] Step 3: Issue the corresponding tripping or alarm command, and record the fault occurrence point and update it to the fault information set F in order of time points. type(i) ; Execute step 4;

[0046] Step 4: Fault information set F type(i) Each time the update is performed, the updated fault information set F type(i) Perform a data processing to obtain comprehensive fault information data F type-t , the fault information comprehensive data F type- Input a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again, and generates a corresponding control decision based on the future time point;

[0047] Step 5: If the actual time point at which the corresponding fault point fails again is earlier than the future time point, update the fault information set F based on the actual fault. type(i) , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set F at the future time point type(i) , repeat step 4.

[0048] Furthermore, the data type type includes: current data, illumination data, temperature data, and partial discharge data.

[0049] Dynamic self-test standard S type The creation process is as follows:

[0050] The corresponding self-check standard is selected according to the current data type. type It includes: one or more of the self-checking standards for current data, self-checking standards for illumination data, self-checking standards for temperature data, and self-checking standards for partial discharge data.

[0051] For example, before monitoring, the device CPU and major components in the control module need to be self-checked to obtain corresponding self-check data. If the self-check data meets the self-check standard, it means that the device CPU and major components are in normal operation, and the first display signal is sent out at a predetermined frequency. For example, if an "operation indicator light" is set, the green light of the operation indicator light flashes once per second, and if a "fault light" is set, the fault light is not on. If the self-check data does not meet the self-check standard, it is determined to be an abnormal operation state, and a second display signal is sent out, such as if the flashing frequency of the operation indicator light is not fixed, and the fault light is red. If it is in an abnormal operation state, it is necessary to inspect and repair the components of the device. Under normal operation, m monitoring modules are selected from all monitoring modules based on demand, and a new definition is given as monitoring module Ψ. (n→m) ; Set the monitoring module Ψ (n→m) The working status of each monitoring module in is set to

[0052] In a further embodiment, the fault information set F type(i) Further expressed as: F type(i) ={f1,f2,…,f n}, n is the number of failures that have occurred in the data acquisition device within the data type type from the start to the end of the current time point, f n is the fault information of the nth fault, which includes: the fault occurrence point and the fault type, expressed in the form of f n =w q→t , q represents the type of fault, and w represents the weight of fault q at time point t.

[0053] The fault information comprehensive data F type-t The method of obtaining is as follows: type(i) The elements in are fused in a parallel superposition manner to obtain the fused spectrum {F t}, where F t =f1∪f2∪…∪f n ; Normalize the fused spectrum to get the sum of the spectrum dimensions, F type-t = dim({F t}).

[0054] The fault prediction model is expressed as follows:

[0055] Calculate predicted values Among them, f k ∈F type(i) ;

[0056] The next time point L(·) is a differentiable convex function, and Ω(·) is a regularization penalty term.

[0057] Under the normal operation of the CPU and main components of the above device, based on the actual demand for monitoring the medium-voltage switchgear, such as monitoring several aspects of the medium-voltage switchgear, the set Ψ={ψ1,ψ2,...,ψ n}, where m≤n. (n→m) The working state of each monitoring module in the system is set to be put into use (ie, changed from the exit state to the input state). The monitoring modules put into use detect and obtain the corresponding monitoring data in real time, and transmit the data to the CPU.

[0058] Real-time acquisition monitoring module (n→m) Each monitoring module in the corresponding monitoring module obtains the current data 1 , 2 , ..., ζ m ; respectively 1 , 2 , ..., ζ m The corresponding judgment conditions are analyzed to obtain the analysis results: if all current data meet the judgment conditions, it is judged that there is no fault; otherwise, it is judged that there is a fault, the corresponding instructions are issued, and the issuance time of the corresponding instructions is recorded.

[0059] The current data monitored by each monitoring module put into use is set as: 1 , 2 , ..., ζ m ; Since each monitoring module monitors different objects, the evaluation criteria for the current data are also different. 1 , 2 , ..., ζ m , and compare them with their corresponding judgment conditions to obtain comparison results, and there are m comparison results. If the m comparison results all indicate that the current data meets the judgment conditions, it is determined that the monitored medium-voltage switchgear has no fault; on the contrary, if one or more comparison results do not meet the judgment conditions, it is determined that the monitored medium-voltage switchgear has a fault, and a corresponding instruction is issued, such as tripping or alarming, and the time point when the response instruction is issued is recorded.

[0060] Based on the current data that does not meet the judgment conditions and its corresponding monitoring module, and the issuance time point of the corresponding instruction, a fault information set in order of time points is generated.

[0061] When a fault occurs, the following steps are performed: When a fault is determined, a trip or alarm command is issued, and the time point t at which the trip or alarm command is issued is recorded. q Based on the current data that does not meet the judgment conditions, and the monitoring module corresponding to the current data, and t q, generate a set of fault information in order of time points; and perform corresponding maintenance based on the set of fault information.

[0062] In other words, for the current data that does not meet the judgment conditions, determine the corresponding monitoring module. Generate a fault information set in time order based on the current data that does not meet the judgment conditions, its corresponding monitoring module, and the time point of issuing the response instruction. The information contained in the fault information set represents the approximate time point when the medium-voltage switchgear fails, the location of the fault, and the data of the fault, which is based on this information to facilitate later maintenance.

[0063] Determining whether the self-inspection data meets the self-inspection standards includes the following steps:

[0064] The acquired self-test data includes at least self-test data about the CPU, memory, power supply, input circuit, and output circuit; if the self-test data of the CPU, memory, power supply, input circuit, and output circuit all meet their corresponding self-test conditions, it means that the self-test data meets the self-test standard; otherwise, it means that the self-test data does not meet the self-test standard.

[0065] The self-test data of the CPU, memory, power supply, input circuit, and output circuit and their corresponding self-test conditions are as follows:

[0066] (1) The self-test data of the device CPU includes instruction function self-test data and interrupt system self-test data: the operation instructions, transfer instructions, jump instructions, and logic instructions are combined into an instruction block with operation functions, and the result of the instruction block is used to judge the instruction function of the CPU. If the result is correct, it means that the instruction function of the CPU is operating normally, otherwise it means that it is not operating normally; an interrupt signal is generated through external hardware, and the normal operation of the interrupt system is checked based on whether the CPU responds to the interrupt. If the CPU responds to the interrupt, it means that the interrupt system is operating normally, otherwise it means that it is not operating normally.

[0067] (2) Memory self-test data: Select a predetermined verification mode, input values ​​into the memory using the predetermined verification mode, and then read them out to observe whether the data changes, thereby determining possible RAM faults.

[0068] (3) Power supply self-test data: By judging the analog signals connected to the remaining two input terminals of the multi-way switch and the analog-to-digital conversion results, it is possible to determine whether the power supply, analog-to-digital converter, multi-way switch, etc. are working properly.

[0069] (4) Input loop self-test data: It is to perform two analog-to-digital conversions on a measured AC signal in one sampling period. If the results of the two analog-to-digital conversions are consistent or the error is within the allowable range, it means that the holding function of the sample and holder is normal.

[0070] (5) Output circuit self-test data: Generally, the duplex output channel is used to send out the trip signal. By taking advantage of the non-corresponding relationship between the two channels, an exclusive OR logic structure is used to detect the damage of a single channel.

[0071] As mentioned above, each monitoring module monitors different objects, obtains different current data, and has different corresponding judgment conditions. Therefore, the judgment process of the current data in each monitoring module is described in detail below.

[0072] The current data includes the following types: current data, illumination data, temperature data, and partial discharge data; correspondingly, the determination conditions include the following types: determination conditions on current data, determination conditions on illumination data, determination conditions on temperature data, and determination conditions on partial discharge data.

[0073] When the current data is current data, the corresponding analysis result acquisition process is as follows:

[0074] The current data at least includes the following data: I l ,Will I l Corresponding to the fixed value O θ , O f , O l A comparative analysis was conducted and the results are as follows:

[0075] (1)

[0076] in, is the current three-phase current monitored in overcurrent section I, overcurrent section II and overcurrent section III; θ takes values ​​of 1, 2, and 3, and 1, 2, and 3 correspond to overcurrent section I, overcurrent section II and overcurrent section III respectively; i takes values ​​of a, b, and c, and a, b, and c correspond to phase a, phase b, and phase c respectively.

[0077] The specific instructions are as follows: When the monitoring module Ψ (n→m) It includes an overcurrent three-stage monitoring module, that is, when the overcurrent three-stage monitoring module is used for monitoring, the current data monitored by the overcurrent three-stage monitoring module is obtained. Included: I a 1 , I b 1 , I c 1 , I a 2 , I b 2 , I c 2 , I a 3 , I b3 , I c 3 .O θ Included: 1 , O 2 , O 3 The analysis and comparison results are as follows:

[0078] The current three-phase current monitored in real time in overcurrent I section: I a 1 , I b 1 , I c 1 , will I a 1 , I b 1 , I c 1 Respectively with the fixed value O 1 Compare and get the comparison result; if I a 1 , I b 1 , I c 1 The current of any phase is greater than the set value O 1 , it is determined that there is a fault and a trip command is issued after a delay. 1 It is 6 to 8 times the rated current.

[0079] The current three-phase current monitored in real time in overcurrent II section: I a 2 , I b 2 , I c 2 , and I a 2 , I b 2 , I c 2 Respectively with the fixed value O 2 Compare and get the comparison result; if I a 2 , I b 2 , I c 2 The current of any phase is greater than the set value O 2 , it is determined that there is a fault and a trip command is issued after a delay. 2 2 to 3 times the rated current.

[0080] The current three-phase current monitored in real time in overcurrent III section: I a 3 , I b3 , I c 3 , and I a 3 , I b 3 , I c 3 Respectively with the fixed value O 3 Compare and get the comparison result; if I a 3 , I b 3 , I c 3 The current of any phase is greater than the set value O 3 , it is determined that there is a fault and a trip command is issued after a delay. 3 1.5 times the rated current.

[0081] Three-stage overcurrent protection refers to current quick-break, time-limited current quick-break and overcurrent protection, which are all protections that react to the increase of current and act accordingly. They cooperate with each other to form a complete set of protection. The difference between the three stages lies mainly in the different principles for selecting the starting current. Among them, quick-break and time-limited quick-break protections are set according to the maximum short-circuit current at a certain point, while overcurrent protection is set according to the maximum load current.

[0082] (2) in, is the current three-phase load current.

[0083] The specific instructions are as follows: When the monitoring module Ψ (n→m) The overload monitoring module is included in the overload monitoring module, that is, when the overload monitoring module is used for monitoring, the current data monitored by the overload monitoring module is obtained. Included: The analysis and comparison results are as follows:

[0084] The overload monitoring module monitors the current three-phase load current in real time: and will With overload current setting value O f Compare and get the comparison result; if The load current of any phase is greater than the overload current setting value O f , it is determined that there is a fault, and a trip command or alarm command is issued after a delay. Among them, the overload current setting value O f 0.8 to 1.2 times the rated current.

[0085] Three-phase overload current protection refers to the thermal capacity of the equipment itself, that is, overload protection.

[0086] (3) Among them, I lis the current zero-sequence current value.

[0087] The specific instructions are as follows: When the monitoring module Ψ (n→m) The zero-sequence overcurrent monitoring module is included, that is, when the zero-sequence overcurrent monitoring module is used for monitoring, the current data I obtained by the zero-sequence overcurrent monitoring module is obtained. l The analysis and comparison results are as follows:

[0088] The current zero-sequence current value I monitored in real time by the zero-sequence overcurrent monitoring module l and compare it with the zero sequence overcurrent setting O l Compare and get the comparison result; if the current zero-sequence current value is greater than the zero-sequence overcurrent setting value, it is determined that there is a fault, and a trip command or alarm command is issued after a delay. l It is 1.5~2A.

[0089] Zero-sequence current protection refers to the use of special zero-sequence current transformers on cable lines to achieve grounding protection. When a ground short circuit occurs in a neutral point direct grounding system, a large zero-sequence current will be generated. The protection formed by the zero-sequence current component can be used as a main ground short circuit protection. Zero-sequence overcurrent protection does not respond to three-phase and two-phase short circuits, and no zero-sequence component is generated during normal operation and system oscillation, so it has good sensitivity.

[0090] When the current data is illumination data, the corresponding analysis result acquisition process is as follows:

[0091] The illumination data includes the current illumination U, and U is compared with the illumination constant U 0 A comparative analysis was conducted and the results are as follows:

[0092]

[0093] The specific description is as follows: Monitoring module Ψ (n→m) An arc monitoring module is included, and current data monitored by the arc monitoring module is obtained, where the current data is illumination data.

[0094] The current illuminance U monitored in real time by the arc monitoring module is compared with the illuminance constant U 0 Compare and get the comparison result; if the current illuminance U is greater than the illuminance constant U 0 , it is determined that there is a fault and a trip command is issued, otherwise it is determined that there is no fault. 0 It is 7500-8000 lux.

[0095] The arc monitoring module uses a three-way cable type arc sensor for monitoring and is installed on the switch cabinet partition plate or frame. Arc protection tripping mode: In addition to the above judgment mode, there is also the following judgment mode: arc + current.

[0096] When the arc unit transmits the light signal (greater than the predetermined light intensity) from the arc sensor to the main control unit, and the current reaches the starting value at the same time, the arc protection action trips. The above two arc protections are selected based on the actual application scenario.

[0097] When the current data is temperature data, the corresponding analysis result acquisition process is as follows:

[0098] The temperature data includes the current temperature value T, T and the temperature constant T 0 A comparative analysis was conducted and the results are as follows:

[0099]

[0100] The specific description is as follows: Monitoring module Ψ (n→m) A wireless temperature measurement and monitoring module is included, and current data monitored by the wireless temperature measurement and monitoring module is obtained, where the current data includes a current temperature value T.

[0101] The current temperature value T monitored by the wireless temperature monitoring module in real time is compared with the temperature setting value T 0 Compare and get the comparison result; if the current temperature value T is greater than the temperature setting value T 0 , it is determined that there is a fault and a trip command or alarm command is issued; otherwise, it is determined that there is no fault. 0 It is 75~80℃.

[0102] The wireless temperature monitoring unit has extremely high reliability and safety. It uses CT passive wireless temperature sensors, which can be directly installed on each high-voltage switch contact, busbar joint, cable joint, transformer and other electrical joints that are prone to high temperatures, and can monitor the real-time operating temperature data of high-voltage equipment.

[0103] The device can monitor 12 wireless temperature signals, a total of 4 groups; 3 points at the cable head, 3 points at the upper contact of the circuit breaker, 3 points at the lower contact of the circuit breaker, and 3 points at the upper busbar, which can be customized by the user according to the installation location. 4 groups of temperature sensors, if the temperature of one sensor in any group reaches the set value, the device will act (trip or alarm) after a delay.

[0104] S404: When the current data is partial discharge data, the corresponding analysis result acquisition process is as follows:

[0105] The partial discharge data includes all partial discharge data J within a predetermined number of AC cycles D UH ; If all partial discharge data J UHF Both are greater than the threshold J bj , it means there is a fault; otherwise, it means there is no fault.

[0106] The specific description is as follows: Monitoring module Ψ (n→m) It includes a partial discharge monitoring module and a partial discharge detection module, and obtains all partial discharge data J within a predetermined number of AC cycles UHF .

[0107] In order to prevent false partial discharge alarms, since partial discharge will always exist once the insulation problem of the switchgear is irreversible, a time parameter is added to the partial discharge alarm processing parameters, that is, each partial discharge value is greater than the alarm threshold value (threshold J) within 30 consecutive (in this embodiment, D is 30, but not limited to 30) AC cycles. bj ) will generate an alarm signal, the threshold J bj :-45~-40dbm.

[0108] The partial discharge monitoring module can detect partial discharge caused by material aging, poor contact, current overload and other factors at the contact points of electrical equipment, and help eliminate potential faults. The partial discharge phenomenon is monitored by partial discharge sensors, which are installed in locations prone to partial discharge, such as switch cabinet cable rooms.

[0109] Example 2

[0110] A switch cabinet comprehensive online monitoring analysis and fault prediction system, used to implement the comprehensive online monitoring analysis and fault prediction method as described above, comprising:

[0111] The first module is configured to configure various types of data acquisition devices D according to the functional requirements of the switch cabinet. type(i) , type(i) represents the number of the data acquisition device in the data type type;

[0112] The second module is configured to create a dynamic self-checking standard S based on the data type type type , the data acquisition device D type(i) Real-time data collected type(i) Input dynamic self-test standard S type , determine whether there is a fault; if there is a fault, execute step 3;

[0113] The third module is configured to issue a corresponding tripping or alarm command and record the fault occurrence point updated to the fault information set F in order of time points. type(i) ; Execute step 4;

[0114] The fourth module is set to the fault information set F type(i) Each time the update is performed, the updated fault information set F type(i) Perform a data processing to obtain comprehensive fault information data F type-t , the fault information comprehensive data F type-tInput a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again, and generates a corresponding control decision based on the future time point;

[0115] The fifth module is configured to update the fault information set F based on the actual fault if the actual time point at which the corresponding fault point fails again is earlier than the future time point. type(i) , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set F at the future time point type(i) , repeat step 4.

Claims

1. A comprehensive online monitoring, analysis and fault prediction method for a switch cabinet, characterized in that: The following steps are involved: Step 1: Configure various types of data acquisition devices according to the functional requirements of the switch cabinet , Indicates the number of the data acquisition device in the data type type; Pre-configured control modules and their connections monitoring modules, any monitoring module includes at least the following working states: input monitoring state φ x , and exit monitoring stateφ y ; Based on the requirements, m monitoring modules are selected from all monitoring modules and given a new definition as monitoring module ; The monitoring module The working state of each monitoring module in is set to φ x ; Real-time acquisition monitoring module Each monitoring module in the corresponding monitoring module obtains the current data 1 , 2 , ..., ζ m ; Since each monitoring module monitors different objects, the evaluation criteria for the current data are also different, so execute step 2; Step 2: Create a dynamic self-check standard based on the data type type , the data acquisition device Real-time data collected Enter dynamic self-test standards , determine whether there is a fault; if there is a fault, execute step 3; Step 3: Issue the corresponding tripping or alarm command, and record the fault occurrence point and update it to the fault information set in time order. ; Execute step 4; Step 4: Fault Information Collection Each time the update is performed, the updated fault information collection Perform a data processing to obtain comprehensive data of fault information , the fault information comprehensive data Input a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again and generates a corresponding control decision based on the future time point; Step 5: If the actual time point at which the corresponding fault point fails again is earlier than the future time point, update the fault information set based on the actual fault , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set at the future time point , repeat step 4; The fault information comprehensive data The method of obtaining is as follows: The elements in are fused in a parallel superposition manner to obtain the fused spectrum ,in, ; Normalize the fused spectrum to get the sum of the spectrum dimensions, ; The fault information set Further expressed as: , n is the number of failures that have occurred in the data acquisition device of data type type from the start to the end of the current time point, The fault information of the nth fault; The fault prediction model is expressed as follows: Calculate predicted values : ;in, ; The next time point : ; is a differentiable convex function, is the regularization penalty term.

2. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 1, characterized in that: The data type type includes: current data, illumination data, temperature data, and partial discharge data.

3. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 1, characterized in that: The fault information includes: fault occurrence point and fault type, expressed in the form of , q represents the type of fault, and w represents the weight of fault q at time point t.

4. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 2, characterized in that: The dynamic self-check standard The creation process is as follows: The corresponding self-check standard is selected according to the current data type. It includes: one or more of the self-checking standards for current data, self-checking standards for illumination data, self-checking standards for temperature data, and self-checking standards for partial discharge data.

5. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 4, characterized in that: When the current data is current data, the corresponding judgment process is as follows: The current data at least includes the following data: , , ,Will , , Correspondence and fixed value , , A comparative analysis was conducted and the results are as follows: ； in, It is the current three-phase current monitored in overcurrent section I, overcurrent section II and overcurrent section III; They correspond to overcurrent section I, overcurrent section II, or overcurrent section III, respectively, and i corresponds to phase a, phase b, or phase c; ； in, is the current three-phase load current; ； in, is the current zero-sequence current value.

6. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 4, characterized in that: When the current data is illumination data, the corresponding judgment process is as follows: the illumination data includes the current illumination ,Will With the constant value of illuminance A comparative analysis was conducted and the results are as follows: 。 7. A switch cabinet comprehensive online monitoring analysis and fault prediction method according to claim 4, characterized in that: When the previous data is temperature data, the corresponding judgment process is as follows: the temperature data includes the current temperature value ,Will With temperature setting A comparative analysis was conducted and the results are as follows: 。 8. A switch cabinet comprehensive online monitoring analysis and fault prediction system, used to implement the comprehensive online monitoring analysis and fault prediction method as claimed in any one of claims 1 to 7, characterized in that: include: The first module is configured to configure various types of data acquisition devices according to the functional requirements of the switch cabinet , Indicates the number of the data acquisition device in the data type type; The second module is configured to create a dynamic self-checking standard based on the data type type , the data acquisition device Real-time data collected Enter dynamic self-test standards , determine whether there is a fault; If there is a fault, proceed to step 3; The third module is configured to issue corresponding tripping or alarm instructions and record the fault occurrence point updated to a set of fault information in time order. ; Execute step 4; The fourth module is set as a fault information collection Each time the update is performed, the updated fault information collection Perform a data processing to obtain comprehensive data of fault information , the fault information comprehensive data Input a fault prediction model, the fault prediction model outputs a future time point when the corresponding fault point fails again, and generates a corresponding control decision based on the future time point; The fifth module is configured to update the fault information set based on the actual fault if the actual time point at which the corresponding fault point fails again is earlier than the future time point. , repeat step 4; if the predicted fault does not occur at the future time point, then update the zero fault information to the fault information set at the future time point , repeat step 4.

Images